Combining neuroscience, psychology and persuasive communication insights in marketing : the effects of social setting, smiling faces and eye-to-eye contact on willingness-to-pay for both symbolic and functional products

(1)

COMBINING NEUROSCIENCE, PSYCHOLOGY AND PERSUASIVE

COMMUNICATION INSIGHTS IN MARKETING:

THE

EFFECTS

OF

SOCIAL

SETTING,

SMILING

FACES

AND

EYE-TO-EYE

CONTACT

ON

WILLINGNESS-TO-PAY

FOR

BOTH

SYMBOLIC

AND

FUNCTIONAL

PRODUCTS

Roy Staal, BSc 5877407

31-07-2014 Master thesis

Master Business Studies, Entrepreneurship & Management in the Creative Industries Amsterdam Business School, University of Amsterdam

Supervisor: B. Kuijken, MSc Second assessor: dr. J.J. Ebbers

(2)

1 Abstract

Prior empirical studies in the advertising research paradigm found effects of perceptual input on attitudes or purchase intention. However, the effects on real consumer behavior were never thoroughly examined. In this paper, the effects affective primes in advertising on willingness-to-pay are investigated. It is predicted that affective content increases willingness-to-pay for both symbolic and functional products. Using an online second price, sealed bid auction platform, two experiments are conducted to assess the impact of the social setting, smiling face and eye-to-eye contact stimuli on willingness-to-pay. The results of both experiments didn’t find support for any of the hypotheses. Implications of the results are discussed to bridge theory with this study’s empirical findings.

Keywords: cheerleader effect; social setting; emotional contagion effect; smiling face; mutual gaze

effect; eye-to-eye contact; symbolic; functional; willingness-to-pay; advertising

(4)

2 Introduction

The final goal most commercial messages share is to persuade target consumers to acquire or adopt a particular product, service or idea (Meyers-Levy & Malaviya, 1999). The big question is: How do commercial messages persuade and influence consumer behavior?

In order to find an answer, a deep dive into the literature on consumer behavior, information processing by consumers, and consumer psychology is required to fully understand what persuasion in advertising is about and how marketers can leverage this knowledge to influence consumer behavior.

This study aims at bridging neurobiological and psychological insights with marketing practice to reduce ‘trial and error’ attempts, and contribute to the advertising research paradigm. More specifically, this study will review and empirically test what affective stimuli are to be used in advertising to increase consumers’ willingness-to-pay. To get there, a better understanding is needed of how people make consumer decisions and how consumer behavior can be influenced.

Stimulus-response theory seems to be a good starting point to better understand and predict consumer behavior (Nord & Peter, 1980). This theory refers to a sequential mental process which describes how stimuli may lead to consumer behavior: consumers digest stimuli, form preferences and attitudes based on feelings and judgments, and develop purchase intent that may lead to an actual purchase (Armitage & Christian, 2003; Zajonc & Markus, 1982). This implicates that people process incoming information consciously, form an opinion, and willingly decide whether or not to engage in courses of action. Other studies, however, found support to question this assumption of the consumer being in control (Bargh & Chartrand, 1999).

Evidence from the nascent field of neuroscience suggests it is more likely that subconscious mental systems, triggered by external stimuli, at least partially, determine what people do (Dunn, Dalgleish, & Lawrence, 2006). Purely rational decision-making, as if people were equipped with unlimited information-processing power, does not exist (Bechara & Damasio, 2005). In other words, when searching for an explanation of what 4

(5)

causes consumer behavior, both internal psychological processes and external influences should be taken into account (De Martino, Kumaran, Seymour, & Dolan, 2006; Nord & Peter, 1980).

Information-processing and, as a result, decision-making are aided by emotions ‘that are elicited during the deliberation of future consequences’ (Naqvi, Shiv, & Bechara, 2006). So, when preferences can be formed without extensive cognitive encoding, this means advertisers can manipulate, or frame, their messages to directly appeal to the consumers’ emotional part of the brain (Zajonc, 1980). Framing, and affective priming, appear to be crucial in guiding choice behavior, since human choices are often based on simplifying heuristics, especially when information is incomplete or overly complex (De Martino et al., 2006; Mograbi & Mograbi, 2012).

Without further discussing its neurobiological basis, human decision-making is subject to affect heuristics underwritten by an emotional system. Moreover, people have the tendency to passively and unintentionally mimic behavior of one’s interaction partners to fit in one’s current social environment, often called mirroring (Chartrand & Bargh, 1999). This is why, for instance, yawning seems to be contagious, and why smiling faces evoke positive feelings toward the smiling people (Lau, 1982; Söderlund & Rosengren, 2003).

Even though exposure to affect-laden stimuli seems to influence evaluative judgments and subsequent behavior, the assumption that this would be the case for all products deserves further investigation. From research on product categorization, ad type and ad processing it was proposed to distinguish between symbolic and functional products (Bhat & Reddy, 1998; Hirschman & Holbrook, 1982; Nelson, 1974). In this paper, it is examined and tested how people process commercial content on both product types and how this impacts behavior.

It was found that three affective stimuli in advertising in particular are among the most effective primers to facilitate positive responses in terms of attitude, intention and consumer decision making (Kuhnen & Knutson, 2011). These stimuli are referred to as the

(6)

social setting stimulus, the smiling face stimulus, and the eye-to-eye contact stimulus. To date, no study has investigated the effect of these stimuli, separately nor combined, on consumer decision-making.

First, it is stipulated that pairing a product with a human model will prompt more positive attitudes compared to a mere product setting (Bandura, 1969; Nord & Peter, 1980). Moreover, individuals are more attractive in a group than in isolation (Langlois & Roggman, 1990; Walker & Vul, 2014). Therefore, a social setting with more than one person accompanying the product is preferred. This is also called the cheerleader effect.

Second, it is brought forward that the smiling face is an effective tool in eliciting positive attitudes toward the smiling person as well as to the firm using the smile appeal (Söderlund & Rosengren, 2003). In other words, the effect of a smile seems to spill over to the observer. It is found that when confronted with someone smiling, an automatic and mostly subconscious effect comes into play: the emotional contagion effect (Wild, Erb, & Bartels, 2001).

Third, making eye contact is a powerful way of establishing a communicative link between humans (Farroni, Csibra, Simion, & Johnson, 2002). This is often labeled as the mutual gaze effect. A mutual gaze demonstrates sincerity, conveys emotion and contributes to the credibility of the sender, especially in the event of non-verbal exchange (Kleinke, 1986; Schore, 2005; Wheeler, Baron, Michell, & Ginsburg, 1979). Thus, examining the effect of the eye-to-eye contact stimulus in advertising might lead to interesting findings.

To date, the effects of these affect-laden stimuli on overt consumer behavior, such as purchasing, remains to be uninvestigated. Most studies focus on the role of affect, feelings and emotions on attitude toward the ad, attitude toward the brand and purchase intent, all of which are assessed by introspection (Brown & Stayman, 1992; Edell & Burke, 1987; Holbrook & Batra, 1987). A measure that bypasses conscious reasoning and predicts decision-making and behavior effectively is needed. Willingness-to-pay is such a

(7)

measure, revealing the ‘maximum amount of financial resources people are willing to give up in exchange for the object being sold’ (Plassmann, O'Doherty, & Rangel, 2007). Moreover, a second price, sealed bid (SPSB) auction appears to be an effective willingness-to-pay elicitation device (Noussair, Robin, & Ruffieux, 2004; Vickrey, 1961). Hence, insights gained from the literature are empirically tested using such an auction system.

This paper combines a thorough literature review with two online SPSB auctions (Vickrey, 1961) to find an answer to the research question:

W hat are the effects of social setting, smiling faces and eye-to-eye contact on willingness-to-pay for both symbolic and functional products?

The structure of this paper is as follows: first, the literature on human decision making, subconscious information processing, priming and product categorization is reviewed. Second, neuroscience and cognitive psychology insights are used to ground and state this study’s hypotheses. Third, method and data collection is described. Fourth, the results of the experiments are presented and will be further discussed, along with the theoretical contribution, managerial implications, limitations and conclusions of this study.

(8)

3 Theoretical framework

Many theories and concepts have been developed in attempts to understand consumer behavior in order to influence it (Peter, Olson, & Grunert, 1999). To throw light on how behavior can be influenced, literature on human choice behavior, information processing and neuropsychology is being reviewed and discussed.

3.1 From stimuli to behavior

Literature on human decision-making has shown that people don’t normally ‘see and buy’. Thus, instigating immediate overt behavior, such as buying, isn’t the advertisers’ first or only objective. Classical stimulus-response theory, therefore, doesn’t tell the whole story.

When exposed to advertising, the consumer process typically starts with taking in the environment: a set of stimuli, or sensory cues, designed to influence consumers’ affect, cognition and, eventually, behavior (Peter et al., 1999). Stimuli that provoke either positive or negative reactions in the individual are ‘value-charged’ (Betsch, Plessner, Schwieren, & Gütig, 2001). These value-charged stimuli don’t impact behavior directly, but rather indirectly by shaping attitudes, preferences and intentions (Armitage & Christian, 2003; Kim, Lim, & Bhargava, 1998; Kuhnen & Knutson, 2011; Zajonc & Markus, 1982).

Attitudes are considered one of the most powerful determinants of behavior and, therefore, are an indispensable construct for understanding and predicting human judgment and decision making (Allport, 1935 in (Betsch et al., 2001), p.242). As a result, most literature on advertising focuses on ad attitudes in measuring advertising effectiveness (Brown & Stayman, 1992).

Whereas attitudes normally are cognition-based, preferences are primarily affect-based behavioral phenomena (Zajonc & Markus, 1982). The interaction between and the relative importance of affect and cognition in the process of attitude formation is still a central theme in attitude-behavior research (Kim et al., 1998; Kuhnen & Knutson, 2011). 8

(9)

When evaluating both affect and cognition in advertising, it is found that the affective element must be altered if preferences are to be changed (Zajonc & Markus, 1982). In other words, using affect is more effective in influencing preferences, attitudes and eventually financial decision making (Kuhnen & Knutson, 2011).

In short, it is argued that the relation between external stimuli and overt behavior is mediated by attitudes, preferences and intentions, and that both affect and cognition significantly contribute to the formation of all mediators (Armitage & Christian, 2003; Peter et al., 1999; Zajonc & Markus, 1982).

3.2 Subconscious information processing

In psychological research it was assumed that people are ‘consciously and systematically processing incoming information’ (Bargh & Chartrand, 1999). In the same vein, the role of emotions on decision-making has been underestimated or even ignored (Bechara & Damasio, 2005). However, recent research suggests that emotions play a pivotal role in financial decision making (Bechara & Damasio, 2005; Kuhnen & Knutson, 2011). And, perhaps even more interesting, they do so largely by bypassing our consciousness (Naqvi et al., 2006). In other words, human decision making can be explained by combining traditional economic theory (e.g. simple cost-benefit analysis), and neurobiological insights, stating that ‘emotions are a major factor in the interaction between environmental conditions and human decision processes’ (Bechara & Damasio, 2005; Naqvi et al., 2006). These decision processes are based on subconscious information processing. It is shown that a vast majority of everyday decisions and actions are based on heuristics, or mental shortcuts, and internal, subconscious forces (Bargh & Chartrand, 1999; Bechara & Damasio, 2005; Dunn et al., 2006).

Another interesting and relevant discovery contributing to explaining decision-making is based on brain structures associated with empathy: mirror neurons. People are socially wired, meaning people need to be accepted and liked, and therefore tend to imitate each other. Even along the line of evolutionary history, automatic imitation, or

(10)

behavioral assimilation, has proven to be advantageous over an absence of imitation (i.e. in the event of seeing someone flee from danger) (Dijksterhuis & Bargh, 2001). Thus, people have an innate tendency to mirror each other’s behavior, often without realizing it (Meltzoff & Moore, 1977). Furthermore, researchers found that the same brain regions are activated when people perform certain behavior and when they see the same behavior performed by others (Rizzolatti, Fadiga, Gallese, & Fogassi, 1996). As a result, marketers can impact decision making by priming value-charged stimuli that speak directly to our subconscious emotional brain, evoking emotions and triggering behavioral assimilation.

3.3 Framing, priming and social learning

It is established that perceptual inputs are translated automatically into corresponding behavioral outputs (Dijksterhuis & Bargh, 2001). A manifestation of this perception-behavior link is called the chameleon effect, referring to subconscious mimicry of behavior of one’s interaction partners to match their own behavior with that of others in their current social environment (Chartrand & Bargh, 1999). The human nature of unintentionally mimicking other’s behavior provides fertile grounds for marketers. Apparently, ads displaying a celebrity or model using and enjoying a product would enhance the probability among the ad viewers of buying and using this product. In most cases, this ‘commercial chameleon effect’ will work best when combined with the mere exposure effect: the more people see something, the more they like something (Zajonc, 1968; Zajonc & Markus, 1982).

Closely related to the chameleon effect is social learning theory, or vicarious learning (Bandura, 1969; Nord & Peter, 1980). This refers to a process which attempts to influence behavior by observing the actions of others (models) and the consequences of those actions (Nord & Peter, 1980). Whereas the chameleon effect causes automatic mimicry of somatic events, vicarious learning ‘educates’ observers by showing behavior and the rewards or punishments deriving from that behavior. Both theories underpin the possibility of influencing consumer behavior by leveraging these psychological insights.

(11)

The concepts of framing and priming are instruments marketers can use to design commercial messages to generate positive emotions, affect and cognitions (Homer & Yoon, 1992). While framing and priming are conceptually connected, understanding the differences can help marketing practitioners to strategically create persuasive messages.

The concept of framing finds its origin in social sciences, but has been incorporated in cognitive psychology and communication studies as well (Benford & Snow, 2000). To frame is ‘to select aspects of a perceived reality and make them more salient in a communicating vehicle’ (Entman, 1993). In advertising, framing equals setting the stage, guiding how people should understand what is being displayed (Denzin & Keller, 1981; Homer & Yoon, 1992). Moreover, the framing effect refers to the finding that ‘alternatives that describe the same option in different formats lead to different decisions’ (Mograbi & Mograbi, 2012).

Whereas framing sets the stage, priming makes some stimuli more salient than others, or carefully highlights elements that are ‘on stage’. The assumption of priming indicates that a target stimulus can activate previously learned cognitive structures, triggering our mental network of associations (Wang, 2007). It has been shown that exposure to affective content impacts later judgment and decision making, as a result of implicit information processing. This phenomenon is called affective priming (Mograbi & Mograbi, 2012).

To conclude, framing and affectively priming messages are most effective when they ‘speak’ to our emotional, subconscious brain. The chameleon effect –people tend to mirror behavior they see- and vicarious learning theory –people learn from seeing behavior and its positive or negative consequences- provides marketers with insights of how to frame and prime their messages.

3.4 Symbolic versus functional products

Zajonc and Markus found that the role of affect and cognition on persuasive information processing depends on the type of product (Zajonc & Markus, 1982). Studies on product

(12)

categorization propose a “symbolic/functional” distinction among consumer goods as two ends on a product type continuum (Bhat & Reddy, 1998). Others employ similar terms: “hedonic/utilitarian” (Hirschman & Holbrook, 1982), “search/experience” (Nelson, 1974), and “feel/think” (Claeys, Swinnen, & Vanden Abeele, 1995). For this study, the terms ‘symbolic’ and ‘functional’ are being used.

A symbolic brand or product satisfies the need for self-expression, social approval and prestige, and ‘relate to the multisensory, fantasy and emotive aspects of product use’ (Hirschman & Holbrook, 1982). For this type of product practical use is only incidental, whereas functional brands and products fulfill immediate and practical needs (Bhat & Reddy, 1998). Moreover, purely functional products are about reducing, removing or avoiding problems (Claeys et al., 1995).

In advertising, when the focus is on addressing symbolic meaning, desired values and affective content, ads are labeled as being transformational or affect-based (Cutler, Thomas, & Rao, 2000; Puto & Wells, 1984). Informational or cognitive-based advertising is about providing facts, describing product attributes and focuses less on affective content.

This paper zooms in on how affective input influences behavioral output for both product types. Remarkably, a lot of research has focused on the influence of perceptual input on purchase decision, instead of actual behavior, such as purchasing (Dijksterhuis & Bargh, 2001). Since it is established that most decisions are both influenced and formed subconsciously, behavior cannot adequately be assessed by introspection. In other words, people often don’t do what they say they will do (Betsch et al., 2001). Hence, purchase intention or even attitude toward the ad would fall short in predicting behavior. Instead, a measure representing behavioral output, that largely bypasses cognitive reasoning, is needed. Willingness-to-pay was found to provide for an effective proxy for real consumer behavior (Noussair et al., 2004).

(13)

In the field of advertising, three affective stimuli in particular are commonly used yet rarely thoroughly studied for marketing purposes: the social setting stimulus, the smiling face stimulus, and the eye-to-eye contact stimulus. These stimuli have received their names from the research field of neuroscience, psychology and even pediatrics. This paper elucidates the origins of these stimuli in order to transform them into testable effects, and attempts to convert scattered research paradigms to valuable marketing insights.

3.5 Cheerleader effect

In advertising, there are roughly two ways of presenting a consumer product: (1) the product in a certain (usage) environment, but without any interaction with human models, or (2) the product accompanied by one or more models, often highlighting its usage experience. The former ad type is in most cases rather informational, whereas the latter automatically evokes more feelings since people are socially wired, as discussed earlier. This social setting, in contrast with a product setting, can also facilitate vicarious learning and mirroring when the ad is designed to convey the product’s experience and its benefits. In addition, it is found that a social setting with more than one model accompanying the product is preferred (Walker & Vul, 2014). People seem more attractive in a group than in isolation. This is named the cheerleader effect. Moreover, average faces were judged as being more attractive, because of evolutionary pressures that favor faces close to the mean of the population (Langlois & Roggman, 1990). In addition, individual faces seem more attractive when presented in a group (Walker & Vul, 2014). Considering the afore-mentioned role of affect on subconscious information-processing and decision-making, it is predicted that the positive affect deriving from a social setting will spill over from an ad to people’s judgment and behavior. It, therefore, shouldn’t matter whether the product is rather symbolic or functional. The following hypothesis is proposed:

H1: Consumers are willing to pay more for both symbolic and functional products when

presented in a group setting as opposed to a product setting.

(14)

3.6 Emotional contagion effect

A social setting may induce subconscious liking and attraction toward the product, a smiling group setting is found to be an even more effective instrument for advertisers to use. People have an innate and universal tendency to smile (Kraut & Johnston, 1979). Reactions to smiling have a prewired neural basis, and are mostly unconsciously evoked (Dimberg, Thunberg, & Elmehed, 2000; Wild et al., 2001). The smiling face stimulus is considered to be a rewarding social signal, enhancing perceived trustworthiness, familiarity, attractiveness, kindness, intelligence and likability (Lau, 1982; Tsukiura & Cabeza, 2008). In fact, smiling is one of the most powerful non-verbal cues in communicating emotion, eliciting favorable attitudes toward the people smiling (Otta, Lira, Delevati, Cesar, & Pires, 1994; Otta, Abrosio, & Hoshino, 1996; Söderlund & Rosengren, 2003).

Most important, however, is the fact that people are susceptible to emotional contagion (Wild et al., 2001). Seeing emotions of others trigger an automatic mechanism to converge emotionally; better known as empathy (Dijksterhuis & Bargh, 2001). This social mechanism already starts at the infancy stage, where babies imitate facial expressions and gestures. This suggests an innate ability to learn from and connect with other human beings (Dijksterhuis & Bargh, 2001; Meltzoff & Moore, 1977). While it is well known that powerful positive emotions can influence behavior, there is evidence that even moderate or low levels of positive affect can alter decision making (Isen, 1984). Hence, a modest but sincere grin would be as effective as enthusiastic laughter in leveraging affect. Although in advertising typical functional products are mostly marketed with informational content, this study suggest that all product types benefit from using emotional contagion in terms of willingness-to-pay. Hence, it is proposed:

H2a: Consumers are willing to pay more for both symbolic and functional products when

presented in a group setting along with the smiling face stimulus as opposed to a non-smiling group setting.

H2b: Consumers are willing to pay more for both symbolic and functional products when

presented in a group setting along with the smiling face stimulus as opposed to a product setting.

(15)

3.7 Mutual gaze effect

When two people looking at each other’s faces and, in most cases, at each other’s eyes, this is referred to as a mutual gaze (Kleinke, 1986). According to many recent studies, making eye contact is the most powerful way of establishing a communicative link between humans (Farroni et al., 2002). To sum up, a mutual gaze conveys emotion and sincerity, demonstrates self-esteem, intelligence and competence; and enhances credibility (Wheeler et al., 1979). Moreover, a gaze provides information, regulates interaction, expresses intimacy and enables people to create an empathic bond with one another (Dijksterhuis & Bargh, 2001; Kleinke, 1986).

As with the smiling face stimulus, the impact of eye-to-eye contact starts at the infancy stage. Babies prefer to look at faces that engage them in mutual gaze, instinctively knowing that affect-laden face-to-face interactions with others convey significant social and cognitive information (Farroni et al., 2002; Schore, 2005). Specifically, in mutual gaze, mother and child engage in subconscious facial communications, increasing levels of joy and excitement (Schore, 2005). A vital skill for social development is to understand that facial expression reflect internal states of social partners, also referred to as ‘theory of mind’ (Farroni et al., 2002).

Thus, a mutual gaze would positively influence consumer behavior and also strengthens the cheerleader effect. Because of the innate susceptibility to mutual gaze, it is proposed that this effect would increase willingness-to-pay for both symbolic and functional products. Hence, the following hypotheses are stated:

H3a: Consumers are willing to pay more for both symbolic and functional products when

presented in a group setting along with the eye-to-eye contact stimulus as opposed to a non-eye contact making group setting.

H3b: Consumers are willing to pay more for both symbolic and functional products when

presented in a group setting along with the eye-to-eye contact stimulus as opposed to a product setting.

(16)

3.8 Combining the effects: Affect accumulation

Combining the aforementioned effects (e.g. cheerleader, emotional contagion and mutual gaze), will lead to a synergy of affective stimuli, exerting an influence on the judgmental process and, eventually, outcome (Cutler et al., 2000; Forgas, 1995). Moreover, highly affective perceptual input induces reliance on heuristic processing, a tendency to mirror perceived behavior and optimally facilitates vicarious learning (Nord & Peter, 1980).

As stated earlier, the smiling face stimulus is considered to be a rewarding social signal, as well as a mutual gaze is found to optimally convey emotions and sincerity. Combining these stimuli in a social setting will elicit positive attitudes and feelings toward that setting (Edell & Burke, 1987; Esch et al., 2012). When a product is placed in such a setting, these positive effects are likely to transfer to that product as well. This is also referred to as the halo effect. More specifically, this effect can be seen as a rating bias or a form of heuristic processing, deriving from traditional psychology research (Beckwith, Kassarjian, & Lehmann, 1978). When trying to judge separate traits of someone, people tend to be influenced by and rely on by a general impression of that person (English, 1934 in (Beckwith et al., 1978). Thus, when applying this insight to marketing communication, this would suggest that consumers evaluate a product not merely on its features, but also on its surrounding stimuli. People’s perception of a product can therefore be influenced by manipulating its context. A positive primer, such as the smiling face stimulus, should increase likability, whereas the eye-to-eye contact stimulus enhances this effect through its ability to convey sincerity. In such a context, the likability toward the product would increase due to an affect spill-0ver. Therefore, it is proposed that the accumulation of the effects will lead to the highest willingness-to-pay via subconscious mental processes.

H4a: Consumers have the highest willingness-to-pay for both symbolic and functional

products when presented in the smiling and gazing group setting

Heuristic processing makes it harder to pinpoint why people perform certain behavior. Often, it is as if a justification is needed after a product has been purchased. Indeed, people tend to justify the consumption and purchase of goods afterwards (Okada, 2005).

(17)

When people are asked to motivate a purchase, it is expected that they ascribe a ‘functional buy’ to facts and features. An ad should therefore be convincing. A ‘symbolic purchase’ satisfies certain desires. In this case, an ad should therefore be rather attractive. As consumers act on heuristic information-processing, it is expected that they ascribe their purchasing behavior to factors that are in line with the type of product that is bought. Thus, the purchase of a rather symbolic product must have been triggered by a rather attractive ad, whereas rather functional purchases would likely be justified due to an ad’s ability to convince. As stated earlier, affect accumulated settings would lead to arousal and a higher willingness-to-pay. It is predicted, however, that product type influences the way people interpret and justify their own behavior. Hence:

H4b: For symbolic products, combining the social setting, smiling face and eye-to-eye

contact stimuli will lead to highest perceived ad attractiveness.

H4c: For functional products, combining the social setting, smiling face and eye-to-eye

contact stimuli will lead to highest perceived ad convincingness.

4 Method

From a marketing perspective, it would be valuable to be able to determine the monetary value people would place on goods to better understand individual and consumer behavior. Experimental auctions can circumvent the problem of asking individuals hypothetical questions about intended behavior (Lusk & Shogren, 2007). Instead, they reflect real behavior since people are expected to bid and commit to their reservation price, or willingness-to-pay.

4.1 Design and Stimulus material

Two online second price sealed bid (SPSB), or Vickrey auctions were conducted, using affective stimuli to elicit real buying behavior, or willingness-to-pay (Vickrey, 1961). At this type of auction, subjects (almost) simultaneously submit a bid. The highest bidder wins, but pays a price equal to the second highest bid (Chakravarti et al., 2002; Noussair et al., 2004). This approach is used, because (1) the online setting makes it easily

(18)

distributable, and accessible to participants; (2) biased introspection is bypassed and actual behavior is measured because bids are binding (i.e. real transactions take place); (3) SPSB auctions are found to be incentive-compatible where truthful bidding dominates over- and underbidding (Riley & Samuelson, 1981); (4) bidder anonymity is guaranteed; and finally, (5) bidders can’t be influenced by time pressure, other bids or the sheer number of bids.

The first experiment will measure willingness-to-pay for a rather symbolic product: a Pathé Home giftcard, worth renting four movies. The second experiment auctions a typical functional product: an electric Bestron popcorn maker. These goods are selected because they differ radically in product type, but are, at the same time, consumed in a similar setting, namely at home when watching a movie. This enhances between-experiment comparability as their designs merely diverge at the display of the product. Thus, the accompanying affective stimuli used in both experiments are identical. Moreover, both products are considered to appeal to both men and women of all ages, enhancing the probability of appealing to an optimal amount of participants, while improving the expected generalizability of the results. However, the suitability of the products are assumed, but not pre-tested.

Five different treatments were constructed in order to investigate the impact of the cheerleader, emotional contagion, mutual gaze and affect accumulation effects on willingness-to-pay. Another two treatments are included to control for the impact of the cheerleader (social setting) effect and, hence, are not used for hypothesis testing. Here, instead of displaying a group of people, only one individual transfers the aforementioned affective stimuli. All treatment designs, for both experiments, are included in appendices 9.1 and 9.3. Both experiments are uniform in their design, except for the displayed products and describing text. All seven treatments are summarized in table 1, along with treatment names.

(19)

Table 1 – Treatment overview

# Treatment content Treatment description

1 Mere product Product setting 2 Social setting stimulus Group setting 3 Social setting + smiling faces stimuli Smiling group setting 4 Social setting + eye-to-eye contact stimuli Gazing group setting

5 Social setting + smiling faces + eye-to-eye contact stimuli Smiling, gazing group setting

6 Individual setting + smiling face stimuli Smiling person (control) setting

7 Individual setting + smiling face + eye-to-eye contact _stimuli Smiling, gazing person (control) _setting

4.2 Variables

Dependent measures Three dependent variables were measured in the study: bid amount in eurocents (willingness-to-pay), advertisement attractiveness (adatt) and

advertisement convincingness (adcon). The latter two variables are used to test hypotheses

4b en 4c; willingness-to-pay (wtp) is the dependent variable in all other hypotheses. Independent variables The factor ‘ad setting’ contains seven levels: the five treatments plus two control treatments each implement each represent a different level. These can be found in table 1, under ‘Treatment description’.

Control variables Whereas adatt and adcon are dependent variables in 4b and 4c,

they can be used as control variables for the other hypotheses. In other words, when no significant difference in wtp is found according to a hypothesis, these measures may still point to useful findings regarding to ad attitudes. In addition, adatt and adcon reflect

participants’ conscious judgment, where wtp, as proposed, is constituted mainly by subconscious mental processes. Comparing these findings might be useful in understanding the role of affective stimuli on decision making.

4.3 Participants

The experimental auctions were conducted using a Dutch, online SPSB auction platform, holding ca.3000 voluntarily subscribed panel members, of which between 1500 and 2000 members are found to be active (i.e. actually participates by submitting bids).Its member base represents the Dutch society in terms of gender, age and socio-economic profile.

(20)

Eight hundred and eighty-three panel members (N = 883) participated in the first experiment and were fairly evenly distributed among the seven treatments: 147, 108, 144, 120, 134, 116 and 114 respectively, with a treatment average of 126 participants. Panel members were randomly assigned to one of the treatments, causing the difference of participants between treatments. Among the participants, 488 (55,3 %) were males. As stated earlier, a Pathé Home gift card was chosen as the experimental product, representing a rather symbolic product. The average bid was € 4, 78.

Nine hundred and twenty-four participated in the second experiment (N = 924). The respective distribution is as follows: 127, 136, 132, 137, 130, 119 and 143, leading to a treatment average of 132 participants. Among the participants, 526 (56,9 %) were males. Here, the experimental product, being rather functional, was chosen to be a Bestron popcorn maker. The average bid was € 4, 79.

4.4 Procedure

Once registered (on complete voluntary basis), panel members were invited by email to participate and bid at a live, one-day auction, and are randomly assigned to one of the seven treatments. Panel members were not aware of the existence of the other treatments than the one they are exposed to. Participants were expected to submit (and confirm) a bid equal to their maximum willingness-to-pay, or reservation price, and have been made aware of the legal obligation to pay when their bid turns out to be the highest (Riley & Samuelson, 1981). Before one can submit a bid, one must first comply with the auction terms and conditions, with the most prevalent term being that ‘bids are binding’. When the auction expires, all participants receive an email announcing the winning bid along with the second highest bid. This is because, as stated earlier, the participant with the highest bid pays only a price equal to the second highest bid. After placing a bid, participants are asked to answer five auction-related questions (see appendices 9.2 and 9.4). Data acquired from these questions are primarily used for robustness checks instead

(21)

of being used for hypothesis testing, except for hypotheses 4b and 4c where rated ad convincingness and ad attractiveness function as dependent measures.

4.5 Hypothesis testing

A one-way between-groups analysis of variance (ANOVA) was conducted to explore the impact of the three aforementioned affective stimuli on willingness-to-pay. This type of ANOVA analysis enables computing and comparing mean scores on a continuous variable of three or more groups, reflecting the impact of one independent variable or factor (ad setting), containing seven levels, on the dependent variable (wtp) (Tabachnick & Fidell, 2007). Moreover, to establish the significance of measured differences among treatments, a LSD post-hoc mean comparison is executed. To test hypotheses 4b and 4c, bid amount is replaced by rated adatt and adcon as dependent variables. In addition, robustness checks are

carried out by splitting the data on bid amount, showing wtp per treatment for both upper and lower 50 % of the bids.

5 Results

First, skewness and kurtosis issues are examined. Second, hypotheses are sequentially tested.

5.1 Normality testing

The data for both experiments are not normally distributed, but are positively skewed. However, this shouldn’t be a problem according to Tabachnick & Fidell. They state ‘with reasonably large samples, skewness will not make a substantive difference in the analysis and also the risk of underestimating the variance (kurtosis) is reduced with a large sample of 200+ cases’ (Tabachnick & Fidell, 2007). Moreover, the positively skewed distribution partly reflects the underlying nature of the construct being measured. Considering the fact that people are encouraged to place a bid of zero (cents) when they are not interested in buying the auctioned product, a positively skewed distribution makes sense. In the two 21

(22)

experiments conducted, respectively 281 participants (or 31.8 %) and 292 (31,6 %) placed a zero bid, causing a non-normal, positively skewed distribution. However, these zero bids are dispersed across all treatments. Moreover, the robustness checks will partially circumvent this issue, when looking at the upper 50 % of the bids.

5.2 Hypothesis testing

5.2.1 Test of H1: Cheerleader effect

Hypothesis 1 predicts that for both experiments wtp will be higher when participants are primed with the social setting stimulus compared to a mere product setting.

Symbolic product The ANOVA revealed that people primed with the social setting stimulus are willing to pay more than participants exposed to the product setting. However, the difference is small (means: 519 versus 464, p > .05). In addition, when looking at the top 50 % bids, filtering out (most) zero bids evenly, the difference remains to be insignificant (means: 961 versus 856, p > .05).

Functional product Results of the second experiment suggest the same pattern concerning the impact of the cheerleader effect. Participants only submitted a slightly higher bid in the group setting than in the product setting (means: 520 versus 488, p > .05). Also, when considering only the highest 50 % of the bids, bid amounts of both settings are almost equal (means: 929 versus 911, p >.05). See results in table 2.

Table 2 – Test of H1: results

Variables Setting N Mean (in € cents) Sig. Symbolic

Bid amount Product 147 464 .42

Group 108 519 Bid amount in top 50% Product 72 856 .26 Group 52 961 Functional

Bid amount Product 127 488 .64

Group 136 520 Bid amount in top 50% Product 62 911 .85 Group 67 929 22

(23)

Conclusion The ANOVA analysis on the impact of the cheerleader effect on willingness-to-pay for both experiments didn’t reach statistical significance. Hence, we found no support for H1.

5.2.2 Test of H2: Emotional contagion effect

Hypothesis 2 suggests that both types of product benefit from being paired with the social setting and smiling face stimuli in terms of willingness-to-pay compared to the group setting (H2a) and to the product setting (H2b).

Symbolic product Results on the added smiling face stimulus showed a decrease in wtp compared to the group setting (means: 472 versus 519, p >.05). This is opposite from what is hypothesized (H2a). As opposed to the product setting, the smiling group yielded only a marginally higher wtp (means: 464 versus 472, p > .05). Excluding the lower 50 % bids resulted in similar findings. Also adatt and adcon for the smiling group setting was rated the

lowest compared to all other settings, indicating a counterintuitive impact of the smiling face stimulus. Interestingly, participants primed with the smiling individual (control) setting were found to have the lowest wtp of all settings, with a significant difference between the group setting (means: 360 versus 519, p < .05).

Functional product In line with the emotional contagion hypothesis, wtp in the smiling group setting was higher compared the group (H2a) and product setting (H2b), although no significance was found (p > .05). Zooming in on the highest bids yielded similar findings. See table 3 for an overview of the results. A noteworthy finding is that willing to pay for functional products has found to be highest when participants are primed with the smiling group setting, although this treatment was found to be least attractive and convincing.

(24)

Variables Setting N Mean (in € cents) Sig.

Symbolic

Bid amount Group 108 519 .50 Smiling group 144 472 Bid amount in top 50% Group 52 961 .45 Smiling group 70 890

Bid amount Product 147 464 .90 Smiling group 144 472 Bid amount in top 50% Product 72 856 .70 Smiling group 70 890 Functional

Bid amount Group 136 520 .43 Smiling group 132 575 Bid amount in top 50% Group 67 929 .17 Smiling group 64 1057

Bid amount Product 127 488 .21 Smiling group 132 575 Bid amount in top 50% Product 62 911 .12 Smiling group 64 1057

Conclusion For symbolic products, the emotional contagion effect had a detrimental impact on wtp. Willingness-to-pay for functional products was, however, highest in the smiling group setting, yet without statistical significance. Thus, both H2a and H2b are not supported.

5.2.3 Test of H3: Mutual gaze effect

Hypothesis 3 postulates that willingness-to-pay for both symbolic and functional products will be higher when participants are exposed to the gazing group setting compared to the group setting (H3a) and to the product setting (H3b).

Symbolic product Priming a gazing group setting increases wtp compared to a product setting (means: 507 versus 464, p > .05). Difference in wtp increases as the lowest 50 % bids are excluded (means: 964 versus 856, p > .05). Therefore, the mutual effect findings are in agreement with H3b, but no support is found since the difference is not significant. Moreover, the added eye-to-eye contact stimulus lowers wtp in comparison with the non-eye contact making group setting (means: 507 versus 519, p > .05). When looking at the highest bids, these settings elicit an almost equal wtp (964 versus 961, p > .05). This contradicts the hypothesized effect in H3a. Self-admitted purchase intention was the

(25)

highest under the gazing group setting, yet the group setting yielded a higher wtp. Table 4 provides an overview of the results.

Functional product Willingness-to-pay is found to be lower in the gazing group setting compared to both the group setting (means: 478 versus 520, p > .05) and the product setting (means: 478 versus 488, p > .05), disconfirming H3a and H3b. Similar findings are reported among the highest 50 % bids. This finding is particularly interesting since perceived adatt and adcon are higher for the gazing group setting than for all other settings

(except the control settings). The gazing group treatment is even found to be significantly more attractive than the smiling group treatment (means: 3.49 versus 3.11, p = .01).

Variables Setting N Mean (in € cents) Sig.

Symbolic

Bid amount Group 108 519 .87 Gazing group 120 507 Bid amount in top 50% Group 52 961 .97 Gazing group 57 964

Bid amount Product 147 464 .51 Gazing group 120 507 Bid amount in top 50% Product 72 856 .23 Gazing group 57 964 Functional

Bid amount Group 136 520 .54 Gazing group 137 478 Bid amount in top 50% Group 67 929 .76 Gazing group 67 900

Bid amount Product 127 488 .90 Gazing group 137 478 Bid amount in top 50% Product 62 911 .91 Gazing group 67 900

Conclusion ANOVA analysis showed that the eye-to-eye contact isn’t effective in increasing wtp for both symbolic and functional products. In fact, the mutual gaze effect in most cases reduces willingness-to-pay. Hence, no support was found for H3a and H3b.

5.2.4 Test of H4: Affect accumulation

Hypothesis 4 predicts that the setting in which the three aforementioned stimuli are combined will yield the highest wtp for both products (H4a). According to H4b, affect accumulation would lead to the highest perceived ad attractiveness for symbolic products.

(26)

H4b states that this setting will yield the highest perceived ad convincingness for functional products.

Symbolic product This study showed remarkable results among the top 50 % bids regarding the effects of affect accumulation on wtp. In contrast with H4a, it is found that wtp for the smiling and gazing group setting was among the lowest, even slightly lower than the product setting where no affective stimuli are incorporated at all (means: 842 versus 856, p > .05). Remarkably, the smiling and gazing individual (control) setting yielded the highest wtp (M = 1046), yet the smiling individual (control) setting elicited the lowest wtp (M = 725), reaching statistical significance (p < .01). Also, the smiling and gazing individual (control) setting differed significantly compared to the smiling and gazing group setting (means: 1046 versus 842, p < .05).

In addition, the smiling and gazing settings (both individual and group) were found to be equally attractive (means: 3.02 versus 3.13, p > .05). In fact, even the product setting was rated higher on attractiveness (M = 3.17).

Functional product Contradicting H4a, wtp (highest 50 % bids) was relatively low when participants were primed with smiling and gazing settings (mean group setting: 858 and mean individual setting: 745). Wtp for the smiling group was even significantly higher compared to the smiling and gazing group setting (means: 858 versus 1057, p < .05). See table 5 with respect to the findings on H4a.

In contrast with H4c, Adcon was found to be lowest (M=2.87) for the smiling and

gazing group setting. Interestingly, the similar control setting was rated highest on perceived convincingness (M = 3.17).

(27)

Table 5 – Test of H4a: results for top 50% bids

(I) Setting (J) Setting (I-J) Mean difference Sig.

Symbolic Smiling and gazing _group

Product -14 .87

Group -118 .21

Smiling group -48 .59 Gazing group -122 .19 Smiling person 117 .21 Smiling and gazing person -204 .03

Functional Smiling and gazing

group Product -53 .58 Group -70 .45 Smiling group -199 .04 Gazing group -42 .65 Smiling person -43 .65 Smiling and gazing person 113 .22

Conclusion The smiling and gazing group setting yielded for both experiments among the lowest wtp measures. Moreover, whereas the smiling and gazing individual (control) setting elicited the highest wtp in the first experiment, the same setting yielded the lowest wtp in the second one. Overall, the results contradict H4a. Perceived adatt (H4b)

and adcon (H4c) didn’t reach statistical significance. Thus, H4a, H4b and H4c are not

supported.

5.2.5 Additional findings

Some additional results that were not covered by the hypotheses appear to be worth mentioning. First, for both experiments wtp for either the smiling or the gazing group setting is higher compared to the smiling and gazing group setting. In fact, when looking at the highest 50 % bids of the second (functional) experiment, the smiling group setting significantly increases wtp compared to the smiling and gazing group setting (means: 1057 versus 858, p < .05). Thus, adding the eye-to-eye contact stimulus has a detrimental effect on wtp. However, results of the first (symbolic) experiment state that the gazing group setting elicits a higher wtp than both the smiling and the smiling and gazing group settings. See table 6 for the average willingness-to-pay for both experiments.

(28)

Table 6 – Mean wtp in eurocents for top 50% bids

Setting N Mean Symbolic N Mean Functional

Product 72 856 62 911

Group 52 961 67 929

Smiling group 70 890 64 1057

Gazing group 57 964 67 900

Smiling and gazing group 64 842 64 858

Smiling person 56 725 58 902

Smiling and gazing person 55 1046 70 745

Total N / Average Mean 426 894 452 898

Second, the smiling and gazing person (control) setting yielded the highest wtp for the first experiment, yet the lowest for the second auction. Remarkably, wtp for this setting (first experiment) is significantly higher compared to both the smiling and gazing group setting (means: 1046 versus 842, p < .05) and the smiling person (control) setting (means: 1046 versus 725, p < .01). Third, the effects of participants’ gender, age and education on wtp differed dramatically across the experiments. There was a significant effect of gender on wtp for symbolic products, F (1, 881) = 8.0, p < .01. Female participants were willing to pay more than males (means: 536 versus 431). Also people above 30 and people with a high education submitted, on average, higher bids (without statistical significance). See table 7 for a complete wtp overview. The opposite was true for functional products. Here, wtp among males was higher. Also participants that were not highly educated bid higher than highly educated participants. In addition, there was a significant effect of age on wtp for functional products, F (1, 922) = 26.4, p < 0.1. Participants younger than 30 were willing to pay more than the older subjects (means: 605 versus 409).

Table 7 – Mean wtp in eurocents, comparing gender, age and education

Gender Age Education

Male Female Under 30 Above 30 Medium/low High

Symbolic 431 536 462 488 451 501

Functional 501 450 605 409 501 489

(29)

6 Discussion

In this section the results are being discussed. In order to answer the research question, the literature is being compared with this study’s empirical findings. Also, both academic and practical contributions are stated. Limitations and suggestions for further research will conclude this section.

6.1 General discussion

This study examined the effects of the social setting, smiling face and eye-to-eye contact stimuli on willingness-to-pay for both symbolic and functional products. Additionally, both perceived ad attractiveness and convincingness were considered to help explain these effects. Previous studies investigated the effects of these stimuli on attitudes or purchase intention. However, no study was found to (1) combine these stimuli and measure its effects, (2) to compare these effects between symbolic and functional products, and (3) to measure actual consumer behavior instead of consumer evaluations.

Based on the literature, it was expected that both product types would benefit from being paired with affective stimuli, both separately and combined. Moreover, the combination of the proposed cheerleader, emotional contagion and mutual gaze effect was predicted to positively influence consumer behavior in terms of willingness-to-pay (both product types), perceived ad attractiveness (symbolic product) and perceived ad convincingness (functional product) through affect accumulation. The results, however, provide no support for either of these predictions. There were no significant increases measured when an affective stimulus was added. Instead, for both experiments wtp varies seemingly random across the settings.

First, people were found to be willing to pay more for products, both symbolic and functional, when primed with the social setting stimulus. However, the difference was far from significant. The rather neutral facial expressions in this setting could have conflicted with general beliefs of how an ad should look like. Indeed, even though wtp increases when primed with the group setting, perceived ad attractiveness decreased, therefore 29

(30)

hindering the impact of the halo effect. It could also be that other factors both enhanced wtp while lowering perceived adatt. For instance, research has shown that people tend to

like objects that other people look at, but these cued objects were liked even more when the stimulus face was smiling (Strick, Holland, & van Knippenberg, 2008). Indeed, the group setting simulated a group implicitly looking at an object, enhancing wtp for that product. However, their neutral facial expressions could be interpreted as a medium or low interest in that particular object, leading to a decrease in ad attractiveness.

Second, to explore if the social setting indeed lacked affect, the smiling face stimulus was added. In effect, willingness-to-pay was expected to increase compared to both the (neutral) group setting and the product setting. Instead, wtp was found to be lower in the smiling group setting compared to both the group and product setting, but only for the first experiment. Moreover, participants rated the smiling group setting as being least attractive and convincing. It could be that the smiling face stimulus was too subtle to notice, or even that the smiling faces were interpreted as being insincere or adverse. Research found evidence that the smiling form and head posture (tilted vs. upright) significantly impact attitude toward the stimulus face(s) (Otta et al., 1994). It is even suggested that some smiles are interpreted as being less attractive than a neutral face. Thus, the impact of different smiles and head postures should be taken into account, since these appear to moderate the relation between a smiling face and ad attractiveness. This study did not consider these suggestions, so it could be that the employed setting evoked rather unfavorable evaluations, explaining the negative ad attitudes and lower wtp. Interestingly, participants were willing to pay the most for functional products when primed with the smiling group setting. The difference in wtp between the smiling group and the product setting even approached statistical significance. Thus, for functional products it appears that an averted gaze, combined with a smile, triggers a vicarious response. It would be interesting to further investigate how product type influences willingness-to-pay and ad perception for smiling group settings.

(31)

Third, studies from the fields of sociology and psychology have found that eye-to-eye contact establishes a communicative link between humans, conveying emotions and expressing intimacy (Dijksterhuis & Bargh, 2001; Farroni et al., 2002; Wheeler et al., 1979). However, the effect of a mutual gaze in advertising on willingness-to-pay has never been examined. The results of this study suggest that a mutual gaze had little impact on wtp. In the second experiment, the proposed mutual gaze effect on wtp was not found, instead the opposite is true: making eye contact lowers wtp for functional products. For the first experiment, this setting elicits only a slightly higher wtp than the product or group setting. In general, the results point to a counterintuitive effect of the eye-to-eye contact stimulus on wtp, opposing the predictions of the mutual gaze hypothesis. Several reasons might explain these findings. First, it might be that a mutual gaze without a smile evokes slightly uncomfortable feelings, interpreting the gaze as a boldly stare. Therefore, participants might perceive the gazing group as being obtrusive, rather than appealing. Second, a direct gaze is considered to be component of attractiveness. It is found that eye-to-eye contact with attractive individuals is perceived to be more rewarding than with less attractive individuals (Fink & Penton-Voak, 2002). This finding might suggest that participants did not perceive the models as being attractive. Third, eye color influences consumer responses in terms of ad attitude and purchase intention. Although eye color in this paper’s treatments was hard, if not impossible, to determine, it should be taken into account. It is suggested that consumers are more likely to be influenced when the model’s eye color matches theirs (Simpson, Sturges, & Tanguma, 2008).

Fourth, it is suggested that the combination of the smiling face and eye-to-eye contact stimuli in a social setting would have a synergetic impact on willingness-to-pay (Forgas, 1995). Thus, positive affect would be conveyed more effectively, whereas a perceived stare should turn into welcoming social signal, optimally facilitating the halo effect to occur. However, the results show an opposite pattern. For both symbolic and functional products, it appears the smiling and gazing group setting yields among the lowest wtp means. In other words, adding a mutual gaze to the smiling faces negatively

(32)

impacts wtp. Research suggests that attractiveness determines how eye-to-eye contact is interpreted. Evaluations are enhanced when ‘attractive faces display direct gaze and not attractive faces display averted gaze’ (Strick et al., 2008). So it could be that (some of) the faces used for this study were perceived as ‘not attractive’. Interestingly, for the first experiment the individual equivalent setting yielded the highest wtp. This suggests that not the eye-to-eye contact stimulus determines how much people are willing to pay, but rather the social setting composition. So varying age, gender, ethnicity and amount of models displayed could very well elicit different wtp levels. Future research should therefore take ad composition, model attractiveness and product type into consideration when designing the settings.

To conclude, willingness-to-pay for both product types is found to decrease when accumulating affective stimuli. This contradicts this study’s prediction. Moreover, no hypothesis found statistical support. Therefore, further research is required to elucidate the effects of social setting, smiling face and eye-to-eye contact stimuli on willingness-to-pay in a marketing context. In section 6.4 suggestions for future research are provided.

6.2 Theoretical contribution

First, this study extends the effects of affect-based stimuli on attitude formation and human behavior to a marketing context. More specifically, findings from various research fields, such as neuroscience, cognitive psychology, subconscious information processing, sociology and persuasive communication, are combined in a marketing context to help explore the potential of perceptual input to influence consumer behavior. Second, this paper contributes to the existing literature on the perception-behavior link, because most studies focus on measuring attitudes and intentions as behavioral output, but this study stipulated that people’s prediction of their own behavior is far from accurate. Focusing on a more reliable measure such as willingness-to-pay, especially when tested in a real market simulation where actual transactions take place, will enhance our understanding of how perceptual input influences behavior. Third, previous studies did not consider the

(33)

importance of product type in the relation between affective stimuli and consumer behavior. This study distinguished between symbolic and functional products and found that the smiling face stimulus is far more effective in increasing wtp compared to the eye-to-eye contact stimulus, but only for functional products. In the case of symbolic products, adding the eye-to-eye contact stimulus increased wtp significantly, but only when presented in a single model setting. Fourth and finally, additional findings suggest that gender moderates the impact on wtp for symbolic products, whereas age is found to significantly impact wtp for functional products. Further research is required to further explore the implications of these findings.

6.3 Managerial implications

Since none of the hypotheses found support, the practical implications of this study are limited. However, the results did point out some interesting suggestions for marketing practitioners to consider. When marketing rather symbolic products, the product should be paired with a smiling model who also makes eye contact with its audience. Willingness-to-pay for this setting is found to be highest. Also, people perceive this setting being most attractive and convincing. Especially the eye-to-eye contact stimulus should be employed, since wtp would plummet without it. Also, adding models to this setting would only decrease willingness-to-pay.

However, when marketing rather functional products, a different approach is required. Here, the eye-to-eye contact stimulus will only decrease wtp. Also, a group of models is preferred over an individual model setting. The highest willingness-to-pay is found when the functional good is placed in a smiling group setting. When this group would establish eye contact with its audience, willingness-to-pay will decrease significantly. Although the smiling group setting would be most beneficial in terms of willingness-to-pay, this setting is also perceived as being least attractive and convincing.

With regard to the proposed effects stated in the hypotheses, no evidence is found for the synergetic effect of the smile and eye-to-eye contact on wtp. Based on the results,

(34)

marketers should focus on other stimuli or stimuli combinations to enhance willingness-to-pay.

6.4 Limitations and future research

There are several limitations with the present investigation. First, parametric tests are used for analysis, while the data for both experiments was found to be positively skewed. Although robustness checks minimized this normality problem, analysis was carried out without pre-testing the assumption of equal variances. Second, two settings are excluded due to an expected shortage of participants. Therefore, this study did not test wtp for the neutral individual setting and the gazing individual setting. Differences in wtp for group versus individual setting, therefore, could not be properly tested. Also, the cheerleader effect and mutual gaze effect could not be fully assessed. To deal with this limitation, the two individual settings that were tested are included as control settings. In effect, no hypothesis is based on these settings. Future research should include these treatments to further explore the effects of social setting on wtp. Third, future research is needed to establish the impact of product type on how to design marketing communications. Also, symbolic and functional products are not necessarily two ends of a one-dimensional scale. Different products can be high or low in both hedonic and utilitarian attributes (Okada, 2005). In the same vein, it is suggested to investigate if the same stimuli trigger different responses for services, or product-service combinations. Fifth, the research method and platform employed for this paper has its shortcomings. To improve participant response rate, bidding on mobile devices (i.e. smart phones) was facilitated. However, ad space was therefore limited to several inches. This might have caused manipulations to be hard, or even impossible, to spot. Affective primes could have been neglected, or even interpreted differently. Therefore, further research should employ different wtp elicitation techniques to cope with this limitation. Sixth, scores on perceived ad attractiveness and ad convincingness are based on a single question after submitting a bid. Also, motivations as to why an ad was perceived a certain way are missing. Now the relation between ad

(35)

attitudes and willingness-to-pay appears to be vague and inconsistent. Further research should investigate this relation to assess to what extent those measures are intertwined. Seventh, the effects of the composition of the social setting would be an interesting and promising topic for future research. Testing mixed model compositions in terms of age, gender, ethnicity, attractiveness and sheer amount could very well lead to different effects on willingness-to-pay, or ad perception. Moreover, varying head postures and eye color could also impact wtp (Simpson et al., 2008; Strick et al., 2008) and should be further tested. Eighth and finally, it must be noted that there is a mismatch between reviewed literature and this study’s empirical findings, since none of the hypotheses found support. However, even without statistical backup, it is urged to further explore the effects of affective stimuli on consumer behavior to enhance (1) psychological insights in a consumer rather than a human context, and (2) marketing effectiveness.

7 Conclusion

Prior research suggests that consumer behavior can be influenced by manipulating perceptual input. Several stimuli have found to be effective in evoking emotions, which could lead to positive attitudes, evaluations and purchase intentions. However, the effects of these stimuli on real consumer behavior were never thoroughly examined.

In this paper, the effects of a social setting, smiling faces and eye-to-eye contact in advertising on willingness-to-pay are investigated. It is predicted these stimuli, especially when combined, increase willingness-to-pay for both symbolic and functional products. Using an online second price, sealed bid auction platform, two experiments are conducted to assess the impact of the social setting, smiling face and eye-to-eye contact stimuli on willingness-to-pay. Moreover, perceived ad attractiveness and ad convincingness are compared to better understand the differences in willingness-to-pay across the different settings. However, the results of both experiments didn’t find support for any of the hypotheses.

(36)

Although the cheerleader, emotional contagion and the mutual gaze effects are established phenomena in psychology and neuroscience, research on consumer behavior falls short in describing how these effects are to be used. Factors such as social setting composition, model attractiveness, head postures, smiling forms and eye color are found to impact perception, attitude and purchase intention. Further research should investigate whether willingness-to-pay can also be influenced by these factors.

To conclude, further examining the potential of affective stimuli to influence real buying behavior instead of attitudes or intentions would be a promising research avenue to take and should be of both academic and practical interest.

Combining neuroscience, psychology and persuasive communication insights in marketing : the effects of social setting, smiling faces and eye-to-eye contact on willingness-to-pay for both symbolic and functional products