OUP user menu

Editors

Darren Dahl (Editor in Chief)Eileen FischerGita JoharVicki Morwitz

3.125
5.003
16 out of 115

The Persuasive Role of Incidental Similarity on Attitudes and Purchase Intentions in a Sales Context

Lan Jiang, Joandrea Hoegg, Darren W. Dahl, Amitava Chattopadhyay
DOI: http://dx.doi.org/10.1086/605364 778-791 First published online: 1 February 2010

Abstract

This study examines the effects of incidental similarity shared between a salesperson and a potential customer. We show that an incidental similarity, such as a shared birthday or birthplace, can result in a more favorable attitude and a higher intention to purchase. We argue and find that the need for connectedness underlies its persuasive effects in an interpersonal context. In addition, we show that the valence of the salesperson’s behavior and the possibility of an extended service relationship moderate the process. When the need for connectedness is mitigated, the positive effects of incidental similarity can be lost or even reversed.

  • Buyer-Seller Interaction
  • Group/Interpersonal Influences
  • Experimental Design and Analysis (ANOVA)

In his Little Red Book of Selling (2005), Jeffrey Gitomer notes that when persuading a customer in a sales situation, “the first thing I do is to establish some kind of rapport that includes finding some common ground” (80). Rapport between consumers and salespeople is typically built through identifying similarities in attitudes and opinions, shared similar experiences, or even something as incidental as shared demographic information. Indeed, the tactic of revealing personal information about personnel has become increasingly popular in sales and customer service contexts. Disney puts the hometown of their employees on their name tags. Hotels such as the Opryland in Nashville and the Hilton in San Diego do the same. The Web sites of many fitness or health care centers have their trainers’ or health care providers’ full biography, even including the high schools they attended, readily accessible for customers. Although such personal information does not provide any specific service to the customer, it helps to create connections and initiate conversations, particularly if the customer shares these similarities (Sommers 2009). Importantly, coincidental matches on such information are not as rare as they sound. For example, the chance for at least two people to have the same birthday is greater than 50% in a group as small as 23 people (McKinney 1966).

The majority of research on similarity examines the role of attitude similarity between individuals (Byrne 1971; Newcomb 1961), and it is well established that people are not only more attracted to others who share similar attitudes but are also more influenced by them (Hendrick and Page 1970; Hodges and Byrne 1972; Reagor and Clore 1970). This research has recently been extended to examine the role of incidental similarity, that is, chance similarities between individuals that provide little relevant information (Burger et al. 2004). For example, a shared birthday with another individual provides no diagnostic information regarding compatibility as business partners. Although it is possible that some people still seek meaning from these similarities, an incidental similarity lies at the lower end of the continuum of informative value, compared to shared attitudes or opinions.

In most instances, logic would indicate that incidental similarities should not play a significant role in social environments. However, existing research has shown that incidental similarity can play a role in social situations and can increase liking, persuasion, and cooperative behavior between individuals (e.g., Burger et al. 2004; Miller, Downs, and Prentice 1998). We extend these initial findings to examine the effects of incidental similarity shared between a salesperson and a potential customer in an actual sales situation. Moreover, we provide an explanation for how and why incidental similarity can have a persuasive influence in an interpersonal context. To do this, we draw from the theoretical framework of a need for belongingness (Baumeister and Leary 1995), where people strive to achieve social connectedness with those around them. This framework is particularly relevant in an interpersonal situation, which is the context of our investigation. We argue and find that the need for social connectedness plays an important role in the persuasive effects produced by incidental similarity.

Across five studies, our research makes three important contributions. First, we establish social connectedness as an underlying mechanism for the effect of incidental similarity. While previous research (e.g., Brendl et al. 2005) has shown that self-esteem enhancement can underlie incidental similarity effects, we show that its influence is overridden by social connectedness in an interpersonal context. Second, we reconcile our social connectedness account with that of self-esteem, revealing situations in which each mechanism is more pronounced. Third, we identify two moderators to the process we outline: the valence of the salesperson’s behavior and the possibility of an extended service relationship with the salesperson. We show that when the salesperson is disliked, the need for social connectedness is mitigated, and, depending on the length of the service relationship, the positive effects of incidental similarity will be attenuated or even reversed.

Incidental Similarity

Research on incidental similarity shared between entities has shown that trivial points of comparison such as birthdates, name letters, fingerprints, and so forth (Burger et al. 2004; Finch and Cialdini 1989; Miller et al. 1998) can influence the attitudes and behaviors of an individual. For example, Miller et al. (1998) had subjects play the Prisoner’s Dilemma game and found that subjects cooperated more often if their birthday was manipulated to be the same as their partners’. Finch and Cialdini (1989) found that subjects’ ratings of a maligned historical figure, Rasputin, were softened when participants believed that they shared the same birthday. Burger et al. (2004) found that a shared birthday, first name, or fingerprint type with a confederate increased compliance with a request made by the confederate (e.g., to donate to a charity). Brendl et al. (2005) found that participants who share a first initial with a brand name are more likely to choose it (e.g., Tonya would be more likely to choose Twix than Layla).

As explanation for this phenomenon, Brendl et al. (2005) focused on self-esteem, suggesting that people possess positive associations about themselves and prefer objects related to the self, as there is a transfer of positive affect from the self to the entity that bears a similarity to the self, incidental or otherwise. They found evidence for this explanation by showing that the optimal condition for the name letter branding effect is under self-threat rather than self-affirmation. While we acknowledge the self-esteem account in explaining the influence of incidental similarity, we argue that in an interpersonal context, other motivations and outcomes that individuals seek to pursue may be more immediate than the need to maintain self-esteem, thereby attenuating its role. We propose that the need for social connectedness (Baumeister and Leary 1995; Lee and Robbins 1995) is such a motivation and has the potential to account for the effects of incidental similarity in an interpersonal context. According to Baumeister and Leary (1995), need for connectedness is fundamental and has been found to drive attitudes and behaviors in a variety of instances (e.g., Argo, Dahl, and Manchanda 2005). Because of the interpersonal nature of the sales experience, where the interaction with another individual is a central and salient component, we expect connectedness to be a primary driver of incidental similarity effects in a sales context. Prior research on incidental similarity effects in interpersonal situations has not tested this potential mechanism (e.g., Burger et al. 2004; Miller et al. 1998). We next outline the scope of our social connectedness theoretical framework and posit specific hypotheses derived from this perspective.

The Need for Connectedness

Social Connectedness and Incidental Similarity

Social connectedness is part of a broader theoretical framework of belongingness, a fundamental human motivation involving the need to form and maintain strong, stable interpersonal relationships (Baumeister and Leary 1995). According to belongingness theory (Baumeister and Leary 1995), humans have an innate need to belong; they are driven to establish and maintain interpersonal bonds. Although there are cultural and individual differences in strength and intensity, the need to belong is universal. It is fundamental in the sense that this motivation is not derived from another motive (see also Kohut 1984).

Lee and Robbins (1995) show that there are multiple aspects of belongingness, including companionship, affiliation, and connectedness. We focus on the social connectedness component of belongingness, “the emotional distance or connectedness between the self and other people” (Lee and Robbins 1995, 239), which we believe is particularly relevant to the interpersonal context we study. Whereas companionship and affiliation involve a need for empathy and support from important individuals (e.g., a spouse, family, parents), connectedness is often seen as a precursor to these more involved forms of belongingness and can be achieved in a broader social context (Newcomb 1990; Timpone 1998). Further, connectedness can be achieved through more subtle means, such as the association created by simple interaction or participation with others (Heider 1958; Lee and Robbins 1995). Thus, we see incidental similarity as a subtle cue that can establish social connectedness in the initial stages of a social relationship.

As a fundamental motivation, the need to connect with others should stimulate goal-directed behaviors. For example, Insko and Wilson (1977) showed that a get-acquainted conversation between two people could lead to strong feelings of association and liking. We argue that an incidental similarity, although trivial, is another type of cue that creates an interpersonal connection, enabling people to satisfy their need for social connectedness. Building a connection has been shown to increase interpersonal attraction (Arkin and Burger 1980). Given that attractive sources are more persuasive, the attraction elicited by the feeling of connectedness should result in greater persuasiveness (Burger et al. 2001; McGuire 1978). Further, it has been shown that individuals give priority to processing information related to people with whom they feel connected (Aron et al. 1991), which could also lead to more effective persuasion.

We propose that an incidental similarity shared between a salesperson and a customer will increase the persuasiveness of the sales pitch, with the feeling of social connectedness between them being the underlying mechanism by which incidental similarity affects attitudes and resulting behaviors (i.e., purchase intentions). More formally:

  • H1: In an interpersonal sales context, an incidental similarity shared between a salesperson and a consumer will increase consumer attitude favorability and purchase intention toward the service or product proffered by the salesperson.

  • H2: Social connectedness felt by the consumer toward the salesperson mediates the relationship between incidental similarity and consumer attitudes.

Limiting the Effect of Incidental Similarity

Valence of Salesperson’s Behavior

Although people have a fundamental need for social connectedness, there are circumstances in which the need for connectedness is mitigated or even rejected. Baumeister and Leary (1995) indicate that to achieve social connectedness and belongingness, interactions with others would be free from conflict and negativity. Forming and maintaining social bonds has survival benefits, such as sharing food and competing for limited resources as a group (Axelrod and Hamilton 1981; Buss 1991). However, when the interactions among group members are negative, dissonance and disliking could turn the survival advantage to disadvantage; therefore, it would not be beneficial to form bonds and connections. Instead, people would want to distance themselves from these types of disliked others. In such instances, an incidental similarity, which creates a sense of association between people, would only aggravate the situation, making it more negative and increasing the desire to disconnect.

Indeed, previous research has shown that people are less willing to interact and connect with similar partners who exhibit negative traits (e.g., Cooper and Jones 1969; Miller et al. 1998; Novak and Lerner 1968). Cooper and Jones (1969) discovered that when a confederate behaved in a belligerent and obnoxious way, participants were more likely to move their opinion away from the confederates if they perceived the confederate as more similar to themselves. Miller et al. (1998) found that people in a Prisoner’s Dilemma game reacted more negatively when betrayed by birthdaymates. Novak and Lerner (1968) showed that people are less willing to interact with a partner who has a history of mental disturbance when they perceive the partner to be similar versus dissimilar to themselves.

Extending these findings, we expect that in a sales context, observing a salesperson with whom an incidental similarity is shared exhibiting negative behavior would decrease the need for connectedness and lead individuals to distance themselves from the salesperson. Since connectedness can underlie the relationship between incidental similarity and attitudes/purchase intentions in this context, distancing should mean negative implications for attitude formation and the resulting purchase intentions.

Possibility of an Extended Service Relationship

Baumeister and Leary (1995) also highlight another feature of the need for belongingness, that is, an enduring and stable interaction. Research has shown that a lack of frequent interactions between parties means that the need for belongingness cannot be fully satisfied (Weiss 1973). It follows that the effects garnered by incidental similarity will best be realized when the opportunity for social connection is ongoing, that is, when a newly established relationship has potential for continuing interaction. In the case of a brief encounter, with no further prospects for contact, the need for connection might be discounted, and the effect of incidental similarity may be mitigated.

Similarly, we expect the negative effects of an incidental similarity with an aversive individual to be mitigated when the social interaction is brief and unlikely to continue. Analogous research in reactance theory has shown that an individual’s resistance to do what is asked is diminished when no future interaction with a social other is anticipated (Andreoli, Worchel, and Folger 1974; Clee and Wicklund 1980). Following the same logic, we expect that the desire to distance oneself from a disliked or negative other, with whom an incidental similarity is shared, would also be reduced when no future interaction is expected.

  • H3A: When an incidental similarity is shared with a salesperson who exhibits negative behaviors, consumer attitude favorability and purchase intention toward the service or product proffered by the salesperson will decrease.

  • H3B: The effect of incidental similarity shared with a salesperson who exhibits negative behaviors will be mitigated when no future interaction is expected between the consumer and the salesperson.

We report five experiments that test our hypotheses. Using a sales situation, Study 1A tests for the effects of incidental similarity in a service context and identifies the role of connectedness in the process (hypotheses 1 and 2). Studies 1B and 2 further validate the role of connectedness by both measuring (study 1B) and directly manipulating (study 2) social connectedness. Study 3 integrates our findings with previous research on the role of another relevant construct, self-esteem, in the incidental similarity effects and tests how these two motives rear their respective heads under different circumstances. Study 4 identifies the importance of the valence of the salesperson’s behavior and the role of expected future interactions with the salesperson in achieving a positive effect of incidental similarity (hypotheses 3A and 3B).

Study 1A

Method

The study was conducted in a marketing behavioral lab. The cover story was that the university recreation center was promoting a personal training program for potential gym members and needed feedback on their program. After signing a consent form, participants were given a brochure containing basic information about the program and a mini-biography of the personal trainer. The short biography described the trainer’s personal information, past experiences, and athletic achievements (see app. A).

The key independent variable, incidental similarity, was manipulated through the mini-biography. Following previous research (Burger et al. 2004; Finch and Cialdini 1989; Miller et al. 1998), we manipulated incidental similarity through shared birthday. Thus, the first line of the biography showed the birth date of the personal trainer. We obtained participants’ birthday information through a general presurvey of the participant pool conducted 2 months before the study. Participants were run in pairs (to maintain a balance between similarity conditions), and for each pair, the trainer’s birthday was manipulated to be the same as one of the participants’. The birthday match was randomly assigned and was inserted in the brochure before the trainer and participants met, eliminating suspicion of the coincidence.

After participants were given sufficient time to read the brochure, a confederate posing as the personal trainer from the recreation center explained the program, following a standardized script. The speech lasted approximately 10 minutes, and participants were encouraged not to interact during this time. The confederate was unaware which participant had the shared birthday, which assured unbiased interaction with both participants and avoided a “contrast effect” for the participants in the nonsimilarity condition. After completing the session, participants filled out a questionnaire containing measures of the program effectiveness (to maintain the cover story) as well as the key dependent variables of interest. Participants were then thanked and debriefed. Fifty-seven university students, aged 18–25, completed the study.

The key dependent variables were intermixed with questions pertaining to the personal training program. Attitude toward the program was measured as an average of two 9-point scale items: “like” and “helpful” (1 = not at all, 9 = very much, $$mathtex$$ $\alpha =.65$$mathtex$$ ). Purchase intentions were measured by two 9-point scale items: “interest in enrollment” and “likelihood to register” (1 = not at all, 9 = very much, $$mathtex$$ $\alpha =.93$$mathtex$$ ). Participants also completed a six-item measure adapted from the Interpersonal Orientation Scale (Swap and Rubin 1983) that measured how connected they felt with the personal trainer on 9-point scales (1 = not at all, 9 = very much, $$mathtex$$ $\alpha =.88$$mathtex$$ ). Representative items included “seem to share the same interest,” “feel connected,” and “willing to talk about my personal life with her.” Gender, age, and participant’s previous experience with fitness programs were also measured and showed no main effects or interactions with the similarity manipulation in this or subsequent studies.

Results and Discussion

Consistent with hypothesis 1, participants who shared the birthday with the personal trainer reported a more favorable attitude toward the program ($$mathtex$$ $M_{\mathrm{sim}\,}=6.92$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=6.37$$mathtex$$ ; $$mathtex$$ $t( 55) =2.03$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). The existence of a shared birthday significantly increased an individual’s intention to enroll in the program ($$mathtex$$ $M_{\mathrm{sim}\,}=5.88$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=5.30$$mathtex$$ ; $$mathtex$$ $t( 55) =2.04$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). Participants who shared the birthday also reported a stronger feeling of connectedness ($$mathtex$$ $M_{\mathrm{sim}\,}=6.33$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=5.54$$mathtex$$ ; $$mathtex$$ $t( 55) =2.39$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ).

Analysis was conducted to test whether connectedness mediates the link between similarity and attitude. Following Baron and Kenny (1986), we show that similarity is a significant predictor of attitude and connectedness. In addition, connectedness is a significant predictor of attitude (t(54) = 3.84, $$mathtex$$ $p< .001$$mathtex$$ ). When both similarity and connectedness are included in the regression model for attitude, the effect of similarity becomes nonsignificant ($$mathtex$$ $t( 53) =.92$$mathtex$$ , NS), while connectedness remains significant (t(53) = 3.37, $$mathtex$$ $p< .005$$mathtex$$ ). A Sobel test of mediation (Sobel 1982) confirmed the mediating role of connectedness ($$mathtex$$ $Z=1.95$$mathtex$$ , $$mathtex$$ $p=.05$$mathtex$$ ). Appendix B presents the path diagram for the mediation analysis.

Discussion

The results of the first study supported our initial hypotheses by confirming the persuasive influence of incidental similarity in a sales context. Supporting hypothesis 1, when participants discovered a shared birthday with the personal trainer, their attitude toward the training program became more positive and their intention to enroll increased. Validating hypothesis 2, we found that the shared similarity increased participants’ feeling of connectedness with the trainer, and this feeling mediated the relationship between incidental similarity and the attitudes formed.

Study 1B

Study 1B seeks to provide further evidence for the articulated process by examining individual differences in social connectedness. We hypothesize that individuals who chronically feel connected to those around them would be positively influenced by an incidental similarity cue, while those who are chronically disconnected would be less affected. In addition study 1B tests the self-esteem account as an additional process involved in the incidental similarity effect observed in study 1A. The current study employs a 2 (incidental similarity: similar vs. nonsimilar) × 2 (self-esteem: threat vs. affirmation) × (chronic connectedness) mixed design, with the first two variables manipulated and chronic connectedness measured.

Method

Sixty-one university students participated in the study. Under the guise of a separate experiment, participants were first asked to complete an unrelated self-knowledge task. In this task, participants were asked either to write down undesirable aspects of who they are and a short description of something that they wish but have not been able to change (“self-threat” condition) or they were asked to write something positive about themselves (“self-affirmation” condition; Jones et al. 2002). The main study was conducted immediately following the writing task and was identical in procedure to that of study 1A.

To test the efficacy of the self-esteem manipulation, we included three 5-point scale items from the state self-esteem measure (Heatherton and Polivy 1991) at the end of the survey. An average of the items (“I feel displeased with myself,” “I feel inferior to others at this moment,” and “I feel like I’m not doing well,” 1 = not at all, 5 = extremely, $$mathtex$$ $\alpha =.84$$mathtex$$ ) constituted the state self-esteem index. The subsequent measures were identical to those used in study 1A with two exceptions. First, attitude toward the program was measured using five 7-point scale items (Good, Like, Helpful, Appealing, and Exciting, −3 = not at all, 3 = very much, $$mathtex$$ $\alpha =.88$$mathtex$$ ). Second, we measured individual differences in social connectedness using Lee and Robbins’s (1995) Social Connectedness Scale (SCS). Eight 6-point scale items (1 = strongly disagree, 6 = strongly agree, $$mathtex$$ $\alpha =.93$$mathtex$$ ) composed the social connectedness index. Representative items included: “I feel distant from people” and “I don’t feel related to anyone,” so a lower score indicated a higher tendency to connect with others.

Results and Discussion

Manipulation Check

We employed a linear regression on the state self-esteem index with the independent variables incidental similarity, self-esteem, the continuous mean centered measure of social connectedness, and their interaction terms (Aiken and West 1991; see also Fitzsimons 2008). The analysis revealed only a main effect of the self-esteem manipulation ($$mathtex$$ $t( 53) =3.24$$mathtex$$ , $$mathtex$$ $p< .005$$mathtex$$ , $$mathtex$$ $\beta =1.99$$mathtex$$ ). Participants in the self-threat conditions reported a lower state self-esteem than those in the self-affirmation conditions ($$mathtex$$ $M_{\mathrm{threat}\,}=3.88$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{affirm}\,}=4.49$$mathtex$$ ).

Attitude and Intentions

We used multiple regression to examine the interactions between incidental similarity, self-esteem, and the centered individual social connectedness score on attitude and purchase intention (Aiken and West 1991). Differing from previous research, none of the coefficients associated with self-esteem was statistically significant ($$mathtex$$ $t{'}\mathrm{s}\,< 1$$mathtex$$ ).

The coefficient for the incidental similarity by social connectedness score interaction, as predicted, was marginally significant for attitude ($$mathtex$$ $t( 52) =-1.81$$mathtex$$ , $$mathtex$$ $p< .10$$mathtex$$ , $$mathtex$$ $\beta =-.87$$mathtex$$ ) and significant for purchase intention ($$mathtex$$ $t( 53) =-2.30$$mathtex$$ , p < .05, $$mathtex$$ $\beta =-.98$$mathtex$$ ). The pattern of effects suggests that the effect of incidental similarity on attitudes and purchase intention is stronger for individuals who chronically feel connected with others. In the shared birthday condition, the social connectedness score predicted attitude (t(29) = −1.65, p < .10, $$mathtex$$ $\beta =-.40$$mathtex$$ ) and intention ($$mathtex$$ $t( 29) =-2.10$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ , $$mathtex$$ $\beta =-.68$$mathtex$$ ). In the nonsimilarity condition, social connectedness did not significantly predict the differences in attitude or intention ($$mathtex$$ $p> .15$$mathtex$$ ). Replicating study 1, the main effect for incidental similarity was marginally significant for attitude ($$mathtex$$ $t( 52) =1.91$$mathtex$$ , $$mathtex$$ $p< .10$$mathtex$$ , $$mathtex$$ $\beta =.88$$mathtex$$ ) and significant for purchase intention ($$mathtex$$ $t( 53) =2.82$$mathtex$$ , $$mathtex$$ $p< .01$$mathtex$$ , $$mathtex$$ $\beta =1.15$$mathtex$$ ). Figure 1 shows the effect of similarity for people with an average SCS score, and a score one standard deviation below/above average.

Figure 1

Attitudes and Purchase Intention for the Personal Training Program as a Function of Incidental Similarity and an Individual’s Social Connectedness Level (Study 1B)

Discussion

Study 1B demonstrates that individuals who have a low social connectedness orientation are least influenced by a shared birthday with a salesperson, whereas individuals with a high social connectedness orientation are influenced strongly by those with whom a birthday is shared. As such, the study provides additional evidence for the role of social connectedness in incidental similarity effects. It appears that although connecting with others is a fundamental human need and a shared birthday can be utilized to build the connection, not every individual is equally likely to be affected. Interestingly, we found no evidence that self-esteem underlies the incidental similarity effects in the sales context we examine. Participants feeling a threat to their self-esteem did not respond to the incidental similarity cue in a differential manner from those experiencing self-affirmation. We return to this issue in study 3.

Studies 1A and 1B support the notion that a need to connect with others contributes to positive effects of incidental similarity. However, in both studies participants’ need to connect was measured rather than manipulated, and in both studies the incidental similarity was a shared birthday. This leaves open the possibility that the influence of social connectedness is limited to a set of very narrow circumstances. Study 2 addresses this concern. We further test the role of social connectedness by directly manipulating an individual’s need to connect and, in addition, increase generalizability by utilizing a different shared similarity: place of birth.

Study 2

As a fundamental human motive, the need to connect is subject to satiation. The satiation hypothesis (Baumeister and Leary 1995) suggests that individuals require a certain degree of social connectedness, but when this need is met, motivation to establish further connections is diminished. Thus, in this study we manipulated the satiation level of the need to connect with others to test whether it would moderate the incidental similarity effects observed previously. To increase generalizability, we used place of birth as the form of incidental similarity in this study. Study 2 employs a 2 (incidental similarity: similar vs. nonsimilar) × 2 (need for connectedness: not satiated vs. satiated) between-subjects design. We expect that when the need to connect is satiated, the effect of incidental similarity will be attenuated.

Method

One hundred undergraduate students at a university in a large West Coast city participated in the study. Each participant’s birthplace had been recorded through an unrelated study conducted several weeks previously. The confederate was of Asian descent and hence could have been born in numerous locations (participants who were born and raised in places in which the shared similarity was unlikely, e.g., Middle East, Eastern Europe, etc., were excluded).

To manipulate connectedness we employed a priming manipulation that has been used in previous research to satiate the need to connect with others (Carvallo and Pelham 2006). The manipulation consisted of a word search task that was conducted under the guise of a different study. Participants were randomly assigned to a satiation or a nonsatiation condition. For participants in the satiation condition, the word search consisted of words related to feelings of connectedness (e.g., accepted, involved, included; Carvallo and Pelham 2006). For the nonsatiation task, participants were asked to search for words unrelated to connectedness but still pleasant (e.g., smile, peace, amuse). To ensure that the priming task would be an appropriate manipulation of social connectedness with our participant pool, we ran a pretest and found that participants completing the connectedness priming task indeed reported a significantly lower need to connect with others than participants completing the neutral priming task ($$mathtex$$ $M_{\mathrm{sat}\,}=4.02$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{not}\,\,\mathrm{sat}\,}=5.06$$mathtex$$ ; $$mathtex$$ $t( 40) =3.42$$mathtex$$ , $$mathtex$$ $p< .001$$mathtex$$ ).

After the priming task, participants were shown a video clip of a personal trainer introducing the training program. The script was the same as the one used in the previous studies. Incidental similarity was manipulated by displaying a short biography of the trainer on the screen, with the first line of the biography indicating the trainer’s place of birth. In the similarity condition, these places were the same as each individual participant’s birthplace. In the nonsimilarity condition the trainer’s place of birth was mismatched with the participants’.

Attitude and purchase intention measures followed those used previously. In addition, we included measures of self-esteem to explore its possible role as a mediator or moderator of the similarity effect. Three sets of self-esteem scales were included in the survey: (1) the 10-item Self Esteem Scale (Rosenberg 1965), (2) the 40-item Self-Esteem Rating Scale (Nugent and Thomas 1993), and (3) the 20-item scale measuring state self-esteem (Heatherton and Polivy 1991). The first two scales measure an individual’s chronic self-esteem, while the third one measures the state self-esteem. Three different self-esteem indices were created by averaging the positive items and the reverse-coded negative items.

Results and Discussion

Attitude and Intentions

A 2 (incidental similarity) × 2 (need for connectedness) ANOVA revealed an interaction effect of incidental similarity and the need for connectedness on participants’ attitudes ($$mathtex$$ $F( 1,\, 80) =4.16$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ) and purchase intentions ($$mathtex$$ $F( 1,\, 80) =4.07$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ; refer to fig. 2). Participants in the not-satiated priming condition reported a more favorable attitude ($$mathtex$$ $M_{\mathrm{sim}\,}=5.50$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.71$$mathtex$$ ; $$mathtex$$ $F( 1,\, 80) =7.07$$mathtex$$ , $$mathtex$$ $p< .01$$mathtex$$ ) and a higher intention to purchase when there was an incidental similarity ($$mathtex$$ $M_{\mathrm{sim}\,}=4.67$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=3.57$$mathtex$$ ; $$mathtex$$ $F( 1,\, 80) =11.33$$mathtex$$ , $$mathtex$$ $p< .005$$mathtex$$ ), replicating study 1A. In contrast, in the satiated condition an incidental similarity did not lead to more positive attitudes ($$mathtex$$ $M_{\mathrm{sim}\,}=4.76$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.83$$mathtex$$ ; $$mathtex$$ $F< 1$$mathtex$$ ) or higher purchase intentions ($$mathtex$$ $M_{\mathrm{sim}\,}=4.00$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=3.83$$mathtex$$ ; $$mathtex$$ $F< 1$$mathtex$$ ).

Figure 2

Attitudes and Purchase Intention for the Personal Training Program as a Function of Incidental Similarity and the Need for Connectedness (Study 2)

Self-Esteem

To test the role of self-esteem we used a series of multiple-regression analyses to test for both main effects of the centered self-esteem scores and interactive effects of incidental similarity, social connectedness, and self-esteem on our key dependent measures. Regression analyses were run with the three different self-esteem indices measured. Consistent with study 1B, none of the coefficients associated with self-esteem was significant ($$mathtex$$ $t{'}\mathrm{s}\,< 1$$mathtex$$ ).

Discussion

Study 2 confirms the role of social connectedness in the incidental similarity effects. We argue that by satiating connectedness, participants were no longer motivated to establish further connection, mitigating the effects of a shared similarity. Considering studies 1B and 2 together, we show through both measurement and manipulation that the need to connect drives the persuasive effect of incidental similarity in an interpersonal sales context. Study 1B examined connectedness from an ability perspective, showing that the incidental similarity loses power for individuals who lack the capacity to receive “appropriate empathy or understanding from peers or society” (Lee and Robbins 1995, 236). Study 2 approached the construct of connectedness from a motivation perspective and showed that the effects of incidental similarity are mitigated when an individual’s need to connect is satiated (Baumeister and Leary 1995).

We again fail to identify a role for self-esteem in our context. Participants with different levels of self-esteem did not respond to the incidental similarity cue in a differential manner. These findings differ from previous research that has shown self-esteem to underlie incidental similarity effects in name letter branding (Brendl et al. 2005). In our next study, we address this discrepancy and propose an integration of our results with previous findings.

Study 3

The studies reported thus far have established the role of connectedness in producing incidental similarity effects but have failed to show a role for self-esteem. However, Brendl et al. (2005) showed that the name letter branding effect was driven by the motive to enhance self-esteem. This discrepancy motivated us to ask whether these two motives coexist in the process or are shown to predominate under different circumstances.

One substantial difference between our studies and those of Brendl et al. (2005) is the nature of the entity that shares the similarity. The name letter was shared between a consumer and a brand, and the birthday or birthplace we studied was shared between two individuals. While self-esteem would likely drive the effects of a similarity between a person and an object, in an interpersonal context its role might be suppressed in favor of a more immediate motive such as social connectedness. If the feeling of connectedness can supersede the need to maintain self-esteem in an interpersonal context, in a noninterpersonal context in which connectedness could not play a role, such as one involving a consumer and a product, rather than a consumer and a salesperson, self-esteem effects should emerge.

Study 3 tests this possibility by investigating circumstances under which self-esteem or social connectedness will emerge as an underlying motive in achieving the incidental similarity effect. We manipulate whether an individual shares a similarity with another individual or with an object. In this study we altered the sales context to a more common and frequently used service, namely, dental care, and created shared similarity either between the customer and the dentist (individual) or between the customer and the clinic (object). We also manipulate self-esteem through either positive or negative bogus feedback in an aptitude test conducted prior to the main study (Arndt and Greenberg 1999; Rhodewalt et al. 1991). We expect that when the similarity is shared with the dental clinic, a threat (but not an affirmation) to self-esteem will lead to positive effects of incidental similarity. In this context, feelings of connectedness will be less affected by the incidental similarity. In contrast, we expect that when the similarity is shared with the dentist (interpersonal context), the feeling of connectedness will increase, resulting in more favorable outcomes regardless of the self-esteem state of the participant.

Method

Study 3 employs a 2 (incidental similarity: similar vs. nonsimilar) × 2 (self-esteem: threat vs. affirmation) × 2 (context: noninterpersonal vs. interpersonal) between-subjects design. Under the guise of a separate study, the self-esteem manipulation was administered first in the form of a student aptitude test, consisting of 14 items from the Graduate Record Examination (GRE). We included only relatively difficult items to make participants uncertain about the accuracy of their answers (Rhodewalt et al. 1991). After completing this test, participants in the self-threat condition were informed that their score ranked in the 11th percentile. In contrast, participants in the self-affirmation condition were told that their score ranked in the 89th percentile.

Participants then were shown a video clip of a dental care service program that is expanding to their area. The beginning of the video showed the dental clinic profile, which contained information about both the dentist and the dental clinic and contained the manipulation of incidental similarity. Consistent with study 2, we chose an Asian confederate to play the role of the dentist. Incidental similarity was manipulated using participants’ birthplace. For participants in the similarity condition, we matched the birthplace of each participant to the birthplace of the dentist or the initial location of the dental clinic, depending on contextual condition. For participants in the nonsimilarity condition, the information was mismatched.

One hundred thirty-three undergraduate students participated in the study. In a suspicion probe, five participants mentioned their suspicion regarding the self-esteem manipulation and were excluded from the subsequent data analysis. We also excluded participants who were born and raised in places in which the shared similarity was unlikely (e.g., Middle East, Eastern Europe, etc.), leaving 113 participants in the analysis.

To test the efficacy of the self-esteem manipulation, we asked participants to rate seven statements from the Heatherton and Polivy (1991) state self-esteem questionnaire, on a 5-point scale. An average of these items created a state self-esteem index ($$mathtex$$ $\alpha =.91$$mathtex$$ ). Attitude toward the dental service was measured using five 7-point scale items: Good, Favorable, Positive, Helpful, and Considerate (1 = not at all, 7 = very much, $$mathtex$$ $\alpha =.94$$mathtex$$ ). Purchase intentions were measured by two 7-point scale items: “consider setting up an appointment” and “likelihood to try” (1 = not at all, 7 = very much, $$mathtex$$ $\alpha =.89$$mathtex$$ ). Feelings of connectedness were measured on a 7-point scale (1 = not at all, 7 = very much).

Results and Discussion

Manipulation Check

A 2 (incidental similarity) × 2 (self-esteem) × 2 (context) ANOVA on the state self-esteem index revealed only a main effect of the self-esteem manipulation ($$mathtex$$ $F( 1,\, 103) =5.32$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). Participants in the self-threat condition reported a lower state self-esteem than those in the self-affirmation condition ($$mathtex$$ $M_{\mathrm{threat}\,}=2.00$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{affirm}\,}=2.51$$mathtex$$ ).

Attitude and Intentions

A 2 × 2 × 2 ANOVA revealed a three-way interaction on participants’ attitudes (F(1,105) = 4.36, p < .05) and purchase intentions (F(1,105) = 5.59, $$mathtex$$ $p< .05$$mathtex$$ ; refer to fig. 3). In the noninterpersonal condition, the interactive effect of incidental similarity and self-esteem was significant for both attitudes ($$mathtex$$ $F( 1,\, 105) =4.92$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ) and purchase intentions (F(1,105) = 5.93, $$mathtex$$ $p< .05$$mathtex$$ ). For threat participants, an incidental similarity led to more favorable attitudes toward the dental service ($$mathtex$$ $M_{\mathrm{sim}\,}=6.45$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.79$$mathtex$$ ; $$mathtex$$ $F( 1,\, 105) =16.01$$mathtex$$ , $$mathtex$$ $p< .001$$mathtex$$ ) and higher purchase intentions ($$mathtex$$ $M_{\mathrm{sim}\,}=6.14$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=3.47$$mathtex$$ ; $$mathtex$$ $F( 1,\, 105) =22.11$$mathtex$$ , $$mathtex$$ $p< .001$$mathtex$$ ). For affirmation participants, the positive effects of the similarity on attitudes and purchase intentions were mitigated (attitude, $$mathtex$$ $M_{\mathrm{sim}\,}=5.48$$mathtex$$ vs. Mnonsim = 5.09; $$mathtex$$ $F< 1$$mathtex$$ ; purchase intention, $$mathtex$$ $M_{\mathrm{sim}\,}=4.59$$mathtex$$ vs. Mnonsim = 3.85; $$mathtex$$ $F( 1,\, 105) =1.87$$mathtex$$ , NS).

Figure 3

Attitudes and Purchase Intention for the Dental Care Program as a Function of Incidental Similarity, Self-Esteem, and the Interpersonal Nature of the Context (Study 3)

In the interpersonal condition, self-esteem did not interact with similarity on attitude or purchase intentions ($$mathtex$$ $F{'}\mathrm{s}\,< 1$$mathtex$$ ). Sharing the birthplace with the dentist increased both attitude ratings ($$mathtex$$ $M_{\mathrm{sim}\,}=6.01$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=5.35$$mathtex$$ ; $$mathtex$$ $F( 1,\, 110) =5.21$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ) and purchase intentions toward the program ($$mathtex$$ $M_{\mathrm{sim}\,}=5.10$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.17$$mathtex$$ ; $$mathtex$$ $F( 1,\, 110) =5.48$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ) regardless of self-esteem. This finding was consistent with the results of the previous studies.

Connectedness and Mediation Analysis

The 2 × 2 × 2 ANOVA on the measure of connectedness, as expected, did not reveal a three-way interaction ($$mathtex$$ $F< 1$$mathtex$$ ). However, it did reveal a two-way interaction between context and similarity ($$mathtex$$ $F( 1,\, 106) =3.95$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). Follow-up contrasts showed that participants in the interpersonal condition reported a stronger feeling of connectedness (Msim = 4.02 vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=2.84$$mathtex$$ ; $$mathtex$$ $F( 1,\, 111) =7.90$$mathtex$$ , $$mathtex$$ $p< .01$$mathtex$$ ) when there was an incidental similarity, but in the noninterpersonal condition, sharing the similarity did not affect their feeling of connectedness ($$mathtex$$ $M_{\mathrm{sim}\,}=3.71$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=3.62$$mathtex$$ ; $$mathtex$$ $F< 1$$mathtex$$ ). Mediation analysis confirmed that the feeling of connectedness mediates the link between similarity and attitude in the interpersonal condition (Sobel $$mathtex$$ $Z=2.22$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). See appendix B for details.

Discussion

The results of study 3 showed that when an incidental similarity is shared with an object, that is, a dental office, it does not affect participants’ feeling of connectedness, and its positive effects on attitudes and purchase intentions are moderated by self-esteem. Individuals whose self-esteem is threatened have a stronger motive to enhance it; thus, they value the service more when they discover a shared similarity. Individuals with higher self-esteem are less motivated to enhance it through praising a similar object; thus, the incidental similarity effects are mitigated. This finding replicates the name letter branding effects in Brendl et al. (2005). In the interpersonal condition, our results showed only positive effects of incidental similarity on attitudes and purchase intentions—regardless of the self-esteem manipulation. Further, we confirmed the mediating role of social connectedness in this context.

It appears that both self-esteem and social connectedness can play a role in defining incidental similarity effects, but the relative influence of one versus the other is a function of the nature of the similarity target and the context. Given that our interest lies in interpersonal sales contexts, the final study focuses on the construct of social connectedness and examines potential boundary conditions on the role connectedness plays in producing incidental similarity effects.

Study 4

Study 4 seeks to test the role of two moderating variables in the process we have identified: the valence of the behavior of the salesperson with whom an incidental similarity is shared and the possibility for an extended service relationship with the salesperson. In accordance with hypothesis 3, we expect that negative behavior on the part of the salesperson will mitigate the need for connectedness and cause individuals to distance themselves from the similar other. As a result, attitude favorability and purchase intentions will both be lower when an incidental similarity with the salesperson exists. However, we expect that this negative outcome will be eliminated if the interaction is a brief encounter instead of a continuing relationship.

Method

Study 4 utilized a 2 (incidental similarity: similar vs. nonsimilar) × 2 (relationship: brief encounter vs. enduring interaction expected) × 2 (valence: nonaversive vs. aversive social behavior) between-subjects design. Similar to study 2, participants were shown a video clip of a personal trainer introducing a training program. The script was the same as the one used in the previous studies. Relationship was manipulated with a 30-second “intro” adapted from the introduction utilized in the previous studies that emphasized whether the presenter would be their trainer or not, if they enrolled in the program. In the brief encounter condition, the intro script indicated that “when you join the program, you will be working with another trainer,” “she will first … ,” “then she will guide you through … ” and “she will be with you the whole time during your workout.” In the enduring interaction expected condition, the script indicated that “when you join the program, I will be your trainer,” and all the “she” words were replaced by “I.”

Between the 30-second intro and the sales pitch, the screen displayed a short biography of the trainer, in which the incidental similarity was manipulated as before. The valence manipulation came at the end of the video clip. In the aversive condition, the trainer rudely complained about the shooting environment to the production staff (off camera). In the nonaversive condition, the trainer was seen thanking the production staff for the shooting experience. Both video segments were portrayed as being accidentally uncut from the video. Throughout the video running time the statement, “This video needs further editing,” was placed at the bottom of the screen to enhance the believability of the manipulation.

One hundred and seventy undergraduate students at a large university participated in the study. In a suspicion probe, 11 participants mentioned their suspicion regarding the behavioral valence manipulation and were excluded from the subsequent analysis.

To test the efficacy of the valence of behavior and relationship manipulations, we included two additional 9-point scale items at the end of the survey: “Do you think the personal trainer is rude/polite?” (1 = rude, 9 = polite) and “How often would you expect to interact with the trainer in the video if you join the program?” (1 = rarely, 9 = all the time). Attitude, purchase intention, and connectedness measures followed those used in study 2.

Results and Discussion

Manipulation Check

A 2 × 2 × 2 ANOVA on the measure of the expected interaction revealed only a main effect of the relationship manipulation (F(1,151) = 6.56, $$mathtex$$ $p< .05$$mathtex$$ ). Participants in the enduring relationship condition reported that they expected to interact more with the trainer than those in the brief encounter condition ($$mathtex$$ $M_{\mathrm{endur}\,}=5.99$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{brief}\,}=4.91$$mathtex$$ ). The valence manipulation check indicated only a main effect for valence (F(1,151) = 30.38, $$mathtex$$ $p< .001$$mathtex$$ ). Participants exposed to the aversive ending reported the trainer to be ruder than those in the nonaversive condition ($$mathtex$$ $M_{\mathrm{avers}\,}=6.02$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonavers}\,}=7.60$$mathtex$$ ).

Attitude and Intentions

A 2 × 2 × 2 ANOVA revealed a three-way interaction on participants’ attitudes (F(1,151) = 7.14, p < .01) and purchase intentions (F(1,151) = 2.80, $$mathtex$$ $p< .10$$mathtex$$ ; refer to fig. 4). In the aversive condition, when participants expected an enduring relationship with the trainer, incidental similarity led to more negative attitudes ($$mathtex$$ $M_{\mathrm{sim}\,}=2.95$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.86$$mathtex$$ ; $$mathtex$$ $F( 1,\, 151) =13.75$$mathtex$$ , $$mathtex$$ $p< .001$$mathtex$$ ) and lower purchase intentions ($$mathtex$$ $M_{\mathrm{sim}\,}=3.09$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.61$$mathtex$$ ; $$mathtex$$ $F( 1,\, 151) =7.23$$mathtex$$ , $$mathtex$$ $p< .01$$mathtex$$ ). However, when no future interaction was expected between the participant and the trainer, the negative effects of the shared birthday were mitigated (attitude, Msim = 4.50 vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.08$$mathtex$$ ; F < 1; purchase intention, Msim = 4.57 vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.27$$mathtex$$ ; F < 1). Interestingly, in the nonaversive condition, sharing a birthday with the trainer increased both attitude ($$mathtex$$ $M_{\mathrm{sim}\,}=4.95$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=4.49$$mathtex$$ ; $$mathtex$$ $F( 1,\, 155) =4.41$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ) and purchase intentions toward the program $$mathtex$$ $( M_{\mathrm{sim}\,}=4.50$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=3.55$$mathtex$$ ; F(1,155) = 9.81, $$mathtex$$ $p< .005$$mathtex$$ ), regardless of whether future interactions with the trainer were expected.

Figure 4

Attitudes and Purchase Intention for the Personal Training Program as a Function of Incidental Similarity, Relationship, and the Valence of the Personal Trainer’s Behavior (Study 5)

Connectedness

A 2 × 2 × 2 ANOVA revealed a marginally significant three-way interaction on participants’ feeling of connectedness ($$mathtex$$ $F( 1,\, 151) =2.61$$mathtex$$ , $$mathtex$$ $p< .10$$mathtex$$ ). In the aversive condition, when participants expected an enduring relationship with the trainer, an incidental similarity led to a lower feeling of connectedness ($$mathtex$$ $M_{\mathrm{sim}\,}=3.24$$mathtex$$ vs. Mnonsim = 5.19; $$mathtex$$ $F( 1,\, 151) =7.77$$mathtex$$ , $$mathtex$$ $p< .01$$mathtex$$ ). When no future interaction was expected between the participant and the trainer, the negative effects of the shared birthday on connectedness were mitigated ($$mathtex$$ $M_{\mathrm{sim}\,}=5.25$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=5.02$$mathtex$$ ; $$mathtex$$ $F< 1$$mathtex$$ ). Following a similar pattern as the attitude and purchase intention measure, in the nonaversive condition, sharing a birthday with the trainer increased the feeling of connectedness (Msim = 6.64 vs. $$mathtex$$ $M_{\mathrm{nonsim}\,}=5.63$$mathtex$$ ; $$mathtex$$ $F( 1,\, 155) =7.19$$mathtex$$ , $$mathtex$$ $p< .01$$mathtex$$ ), regardless of whether future interactions with the trainer were expected. Mediation analysis confirmed the mediating role of social connectedness in producing the identified pattern of effects in the aversive condition (Sobel $$mathtex$$ $Z=1.99$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ) as well as in the nonaversive condition (Sobel $$mathtex$$ $Z=2.56$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). Appendix B presents the path diagram for the mediation analysis.

Discussion

In support of hypothesis 3A, our results showed that when an individual shares an incidental similarity with someone exhibiting an undesirable social behavior (i.e., rudeness), the need to connect with the person is mitigated. Aversive behavior on the part of the similar other caused participants to distance themselves and feel less connected to the other. This decreased feeling of connectedness led to a more negative attitude toward the program and lowered the intention to purchase. Moreover, we show that, consistent with hypothesis 3B, the possibility for an extended relationship moderates the negative effects of incidental similarity. Although sharing the same birthday with an aversive figure is threatening, an individual does not seek to distance themselves from the aversive other as long as no future interactions are expected.

In the nonaversive condition, our results showed only positive effects of incidental similarity—regardless of the nature of the relationship. Baumeister and Leary (1995) proposed that belongingness can be fully satisfied only with frequent and enduring social interactions, although their discussion focused on the companionship and affiliation components of belongingness (Weiss 1973). Our findings focus on the more basic aspect of connectedness and are compatible with other research which shows that feelings of connectedness can be created under minimal relationship conditions (Argo et al. 2005; Insko and Wilson 1977).

Additional post hoc analysis indicated that different processing approaches might partially explain the differing pattern of effects for the aversive and nonaversive conditions. Participants in the aversive condition reported more thoughts than those in the nonaversive condition ($$mathtex$$ $M_{\mathrm{avers}\,}=3.17$$mathtex$$ vs. $$mathtex$$ $M_{\mathrm{nonavers}\,}=2.75$$mathtex$$ ; $$mathtex$$ $F( 1,\, 155) =5.52$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). More relevant to our findings, a 2 (incidental similarity) × 2 (valence) × 2 (relationship) ANOVA on the number of thoughts revealed a two-way interaction between valence and relationship ($$mathtex$$ $F( 1,\, 155) =3.77$$mathtex$$ , $$mathtex$$ $p< .05$$mathtex$$ ). When an enduring relationship was expected with an aversive other the number of cognitive thoughts regarding the service significantly increased (compared to the other three conditions, all $$mathtex$$ $p{'}\mathrm{s}\,< .05$$mathtex$$ ). This suggests that participants thought more deeply about the implications of an aversive other with whom they would have a continuing relationship. Interestingly, when the trainer was nonaversive or the interaction was brief, participants did not process deeply, suggesting more heuristic processing of the incidental similarity cue and the sales experience.

General Discussion

This paper presented five studies to demonstrate how an incidental similarity, that is, a trivial association between individuals, can have a persuasive influence in a sales context. We showed that after a sales promotion for a service program (e.g., personal trainer or dental care), consumers who found out that they shared the same birthday or birthplace with the service representative reported more favorable attitudes toward the program as well as higher purchase intentions. We identified the role of social connectedness underlying this effect through mediation analysis, an individual difference measure, and a direct manipulation. We also investigated boundary conditions for the positive effects of incidental similarity, that is, the valence of the interaction and the length of the relationship. We showed a reversal of the positive effect of incidental similarity when the salesperson exhibited aversive behaviors during the social interactions. However, this negative effect was mitigated when the social interaction was just a brief encounter with no future interactions expected. Interestingly, the possibility of an extended service relationship only moderated the effects of incidental similarity when the interaction was aversive. In nonaversive conditions, an incidental similarity increased attitude favorability and purchase intentions even when the interaction with the similar other was a brief encounter. This finding speaks to the possibility that incidental similarity with a salesperson may lead to positive purchase outcomes even for products that do not involve a future service component.

From a theoretical perspective, our studies distinguish the role of social connectedness from the previously studied construct of self-esteem and elaborate circumstances under which one motive versus the other would be a dominant driver of incidental similarity effects. We demonstrate that both motives can operate as underlying mechanisms for incidental similarity effects, but their relative influence differs depending on the context. Specifically, in interpersonal sales situations, such as those we study, where interaction with the service provider is a key component, social connectedness, rather than self-esteem maintenance, is facilitated. In contrast, when the sales context is not interpersonal and involves only the individual and the product or service, self-esteem seems to exert a larger influence. To our knowledge, this research is the first empirical test of social connectedness as a theoretical explanation for when and why individuals will be affected by an incidental similarity.

Although the results of studies 1B–3 seem to suggest an independent relationship between self-esteem and social connectedness in driving the incidental similarity effects, we do not claim that the two constructs are unrelated. Indeed, social connectedness has been argued to be a factor in reducing susceptibility to low self-esteem (Kohut 1984; Williams and Galliher 2006). However, the null effects associated with self-esteem in our studies suggest that in an interpersonal context the motivation to connect may be primary, thereby mitigating the role of self-esteem. Indeed, our findings provide the impetus for additional research aimed at further delineating the relationship shared between these two constructs. While we speculate that the need for connectedness is primary, what makes social connectedness a more immediate motive than self-esteem? Perhaps in a social setting, connectedness is viewed as a means to achieve self-esteem? Or achieved connectedness simply mitigates the need for self-esteem maintenance? Future research needs to build on our findings to bring additional clarity and a deeper understanding of how incidental similarity affects individuals in a social arena.

From a managerial perspective, our research provides insight into the power of cultivating similarity between consumers and sales agents in situations aimed at getting a person to engage in future activities with the service provider. This paradigm is common to many service encounters, such as those involving financial service providers, lawyers, hair dressers, medical service providers, instructors, mechanics, and so forth. In such situations, particularly when the customer lacks expertise in the area and the service is customized and delivered over time, establishing rapport with a client at an early stage is critical (Crosby, Evans, and Cowles 1990). An incidental similarity can play an important role in this regard.

Prior work in personal selling has examined the interpersonal relationship between buyers and sellers and reports that several personal and emotional factors can play roles in establishing rapport between a salesperson and a customer (Nicholson, Compeau, and Sethi 2001). Such research highlights perceived similarity with a salesperson in terms of shared attitudes, lifestyle, or values as a way to improve liking for and trust of the salesperson, and increase perceptions of salesperson effectiveness. The discovery of similarities on important characteristics of the relationship can increase customer confidence that they will not be misled by the salesperson because they see the world the same way (Nicholson et al. 2001). Our research builds on this work by demonstrating that establishing a connection based merely on incidental similarities may be sufficient to influence consumer decision making. Merely having staff members put their hometown on their name badges can help to initiate conversation and make customers feel that they are special and more familiar (Herbst and Lloyd 2008).

We note, however, that establishing incidental similarities will likely be effective only if the incidental similarity is perceived to be coincidental and genuine. Creating misleading or fake similarities with a customer as a persuasion technique could backfire if the similarities were found to be contrived (Campbell and Kirmani 2000). To mitigate the chances of this outcome, salespeople must be careful not to falsely claim similarities. They should also balance the extent to which they probe for information with the extent to which they share information about themselves. Reciprocal divulgence can lessen feelings of mistrust while still allowing for the discovery of an incidental similarity (Crosby et al. 1990). The shared similarity could also have negative consequences if the salesperson’s subsequent behavior is perceived to be aversive. As our research shows, in such situations, the positive effect of an incidental similarity is not merely attenuated but is actually reversed and becomes a detriment to the sales encounter.

There are numerous opportunities for future research. As noted above, providing additional insight into the underlying mechanisms of connectedness and self-esteem is important. It would also be interesting to examine whether a shared similarity on an attribute where an individual wants to be unique (e.g., a shared first name) or an individual’s chronic need to be unique would lead to different consequences. Another direction for future research could be to study incidental similarity as a continuous rather than a binary variable. As noted, we believe that the incidental similarities we examine lie at the low end of the similarity continuum, but less personally meaningful similarities, such as a shared home street name in different cities or shared final digits in a phone number, could easily exist. We suspect that as the similarities become less personally relevant they will require greater rarity to have the same effects.

Appendix

Figure A1

Biographical Description

Appendix

Figure B1

Path Diagrams for the Mediation Analyses

Note.—Values in parentheses indicate the effects of the similarity or connectedness on attitude when the mediator is included in the model; * indicates significant relationships ($$mathtex$$ $p< .05$$mathtex$$ ) and ** indicates significant relationships ($$mathtex$$ $p< .01$$mathtex$$ ).

References

View Abstract