rs

in

1, 56

ond

f th

vior

avi

r no

stom

ed w

, an

ions

he b

tha

om

& 2010 Elsevier Ltd. All rights reserved.

1. Introduction

1.1. Interpersonal bonds in employeecusto

The interaction between store employonsiderualityr (Bitneonal bos and ta posisiderw

andntactornik,

personal relationships. Within the various social psychology

Contents lists available at ScienceDirect

journal homepage: www.else

Journal of Retailing and

Journal of Retailing and Consumer Services 18 (2011) 381388its effect on the perception of the mimicker are relatively recentangelique.martin@univ-ubs.fr (A. Martin), Gaelle.boulbry@univ-ubs.fr (G. Boulbry).procedures, one of them, the mimicry technique, could have someinterest for managers and sellers to create a positive perception ofthe employees among the customers. Mimicry has long beenstudied by social psychologist. However, its behavioral effect and

0969-6989/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jretconser.2010.11.006

n Corresponding author.

E-mail addresses: celine.jacob@univ-ubs.fr (C. Jacob),

nicolas.gueguen@univ-ubs.fr (N. Gueguen),Thus, it could be interesting for managers to encourage theiremployees to use social psychological procedures that facilitate

1992a) or the restaurant (Hornik, 1992b).Tactile contact is not the sole technique that facilitates inter-sonal bonds in their relationships with the customers such asemployee familiarity with customers, personal connection betweenemployees and customers, or care displayed by employees.

and Zweigenhaft, 1986; Wycoff and Holley, 1990; ErceauGueguen, 2007). It has also been found that employee tactile cowas associatedwith amore positive perception of the store (Hthe development of interpersonal bonds, including encouraging thedevelopment of friendships between employees and customers.Coulter and Coulter (2000) found that as perceived similaritybetween customers and service employees increased, customertrust also increased. Gremler et al. (2001) found that customerpositive word of mouth increased when employees used interper-

potential customers can also lead to an increase in product salesrates, or a greater amount of money spent (Gueguen et al., 2007;Gueguen and Jacob, 2006; Hornik, 1992a; Kaufman and Mahoney,1999; Smith et al., 1982). Further studies found that the professionalqualities of the seller or the restaurant employee were evaluatedmore positively (Crusco and Wetzel, 1984; Hornik, 1992b; Stephenemployees and their customers is ccustomers assessments of service qwith the store or the service provide1998). The development of interpersstore to differentiate itself from otherraise customer loyalty and to createtheir store. Thus,managersmight conmer relationships

ees or service industryed an essential part ofand their relationshipr, 1990; Gwinner et al.,nds may be a way for ao increase their sales, totive word of mouth forays they could facilitate

positive interpersonal relationships in their own relationshipswiththe customer. Several studies have found that some of these socialpsychology techniques facilitated interpersonal relationships andincreased the positive perception of the individual who used suchinterpersonal bonds. Some of these studies have shown theefciency of these interpersonal techniques in a selling context.For example, touch is considered as a factor that facilitatesinterpersonal relationships, familiarity and truth. Several studieshave found that the tactile contact of a patron by a server in arestaurant or bar increases tipping (Crusco and Wetzel, 1984;Stephen and Zweigenhaft, 1986; Hornik, 1992b; Lynn et al., 1998;Ebesu Hubbard et al., 2003; Gueguen and Jacob, 2005). TouchingRetail salespeoples mimicry of custome

Celine Jacob a, Nicolas Gueguen a,n, Angelique Marta Universite de Bretagne-Sud, IUT de Vannes, Departement TC, 8, rue Montaigne, BP 56b Universite de Bretagne-Sud, UFR DSEG, Rue de la Loi, 56000 Vannes, France

a r t i c l e i n f o

Available online 22 June 2011

Keywords:

Mimicry

Sellercustomer interaction

Consumer behavior

a b s t r a c t

Developing interpersonal b

and positive perceptions o

verbal and nonverbal beha

and inuenced helping beh

were instructed tomimic, o

On their way out, these cu

that mimicry was associat

during the selling process

found that these evaluat

Experiment 2 conrmed t

These results seem to show

their sellers and their cust: Effects on consumer behaviora, Gaelle Boulbry b

017 Vannes, France

s between employees and customers in selling contexts can increase sales

e employees and the store. Recent studies have found that mimicking the

of strangers enhanced their liking for the individual whomimicked them,

or. An experiment was carried out in a retail setting where four sales clerks

t, some of the verbal expressions andnonverbal behavior of the customers.

ers were asked to evaluate the sales clerks and the store. Results showed

ith a higher sales rate, greater compliance to the sales clerks suggestion

d more positive evaluations of both the sales clerks and the store. It was

mediate the relationship between mimicry and customers behavior.

ehavioral effect of mimicry when a baseline condition was introduced.

tmimicry really helpsmanagers to develop positive relationships between

ers.

vier.com/locate/jretconser

Consumer Services

C. Jacob et al. / Journal of Retailing and Consumer Services 18 (2011) 381388382and the inuence ofmimicry in a real selling context does not exist.The rst objective of this paper was to test the effect of mimicryused by a salesperson on customer behavior and judgment in a realeld setting. The second objective was to test the link betweencustomer behavior toward a salesperson who mimicked him/herand the variation in judgment created by mimicking.

1.2. Mimicry in social relationships

As the saying goes Monkey see, monkey do. This Chameleoneffect (Chartrand and Bargh, 1999) refers to the unconsciousmimicry of postures, facial expressions, mannerisms and otherverbal and nonverbal behaviors. Today, research on this paradigmhas shown that mimicry has the power to inuence social judg-ment and behavior toward the mimicker. Recent experimentalstudies conducted in eld settings show that using mimicry couldinuence compliance towards a mimickers request.

1.3. Mimicry: an automatic behavior

It has been found that posture and mannerisms are routinelymimicked in social interaction (Bernieri, 1988; LaFrance, 1982).Giles and Powesland (1975) found that peoplemimic the accents oftheir counterparts. Speech rate (Webb, 1972) and speech rhythms(Cappella and Panalp, 1981) are also mimicked unconsciously byhuman beings in their social interactions. The contagious effect oflaughter has been found in several studies (Bush et al., 1989;Provine, 1992) and many experiments have found that the use ofcanned laughter causes an audience, in return, to laugh longer,more often, and to rate the humorous material as funnier (Cialdini,2011). Chartrand and Bargh (1999) showed that participants weremore likely to touch their own face when they interacted with aface-touching confederate who was a stranger, than when theyinteracted with a foot-shaking confederate. In the same way, thereverse effect was found when the confederate shook his or herfoot: the participantsweremore likely to shake their own feet thanto touch their own face. When confederates were instructedto smoke in a bar-laboratory, it was found that participantsimitated this smoking behavior (Harakeh et al., 2007). Quigleyand Collins (1999) found that alcohol consumption is inuencedby our counterpart in social interaction and the type of drinkselected, the drinking rate, and the volume of beverage foreach sip is imitated. However, these later studies found that peoplewere not conscious that they mimicked the behavior of theircounterparts.

1.4. Mimicry increases positive evaluation of the mimicker

Research has also found that mimicry is associated with a morepositive evaluation of the mimicker. Chartrand and Bargh (1999,study 2) engaged participants in a taskwith a confederatewhowasinstructed to either mimic the mannerism of the participant, or toexhibit neutral, nondescript mannerisms. Participants who weremimicked by the confederate subsequently reported a highermeanof liking for the confederate, and described their interactionwith him/her as smoother and more harmonious than those whowere notmimicked. This result is congruent with previous work byMaurer and Tindall (1983), who found that when a counselormimicked a clients arm and leg position, this mimicry enhancedthe clients perception of the counselors level of empathycompared to when the counselor did not mimic the client.Interacting in an immersive virtual reality with an embodiedarticial agent mimicking our own behavior is sufcient toinuence the agents rating. In a recent experiment by Bailenson

and Yee (2005), a virtual agent verbally presented a persuasiveargument (a message advocating a campus security policy) to aparticipant. In half of the cases, the virtual agent mimicked theparticipants head movements with a 4-s delay; for another groupof participants, the agentmimicked the prerecordedmovements ofanother participant. After the interaction, the participant indicatedhis/her agreement with the message delivered by the agent andgave his/her impression of the agent. It was found that themimicking virtual agent was more persuasive, and received morepositive trait ratings than non-mimickers.

1.5. Mimicking someone to create afliation and rapport

Rapport and afliation are also associated with mimicry.LaFrance (1979) conducted a longitudinal design survey to explorethe relation between a measure of nonverbal synchrony and self-report indications of rapport with college students. She found thatposture sharing between the instructor and the students waspositively related to rapport. For this author, postural mimicrymay be inuential in establishing rapport. This link betweenafliation and rapport has recently been demonstrated by Lakinand Chartrand (2003). In their experiment, participants performeda rst task in which they were exposed to a priming procedureusing words related, or not, with the concept of afliation (friend,partner, etc.). In a second unrelated task on memory, participantswatched a videotape of a female confederatewhowas touching herface. During this second task, itwas found that the participantswhowere primed with the unconscious concept of afliation mimickedmore favorably the confederate displayed on the video tape thanwhen no afliation priming was used. This seems to show thatafliation and mimicry are related. This effect was recentlyconrmed by two studies by Yabar et al. (2006), who found thatparticipants mimicked more favorably the nonverbal behavior ofmembers of their in-group than members of their out-group. Onceagain, the link between mimicry and the desire for afliation wasattested. The same effect can be found in closer interaction. In arecent study (Gueguen et al., 2007) an experiment was carried outduring real sessions of speed-dating, in order to test the behavioraleffect of mimicry in a courtship context. The young womenconfederates who interacted with men during such sessions wereinstructed to mimic some of the mens verbal expressions andnonverbal behavior, or only verbal expressions, or nonverbalbehavior alone, or no mimicry. Data showed that men evaluatedmore positively both the dating interaction and the womenconfederates when the latter mimicked them. At the end of thedating session, men expressed a greater desire to meet again thewomen-confederates who had mimicked them than the samewomen-confederate when they had not.

1.6. Behavioral effects of mimicry

If mimicry is associated with a greater liking for the mimickersand a greater feeling of afliation, several studies have found thatmimicry leads to enhance pro-social behavior toward the mimick-ers. Van Baaren et al. (2003) found in two experiments thatmimicking the verbal behavior of customers in a restaurantincreased the amount of the tips. In their rst experiment, awaitress was instructed to mimic the verbal behavior of half ofher customers by literally repeating their order. It was found thatthe waitress received signicantly larger tips when she mimickedthe patrons than when she did not. In a second experiment, it wasfound that compared to a baseline condition, mimicry was asso-ciated with a higher rate of tipping customers, and also with largertips. Spontaneous helping behavior is also affected bymimicry. VanBaaren et al. (2004) mimicked the posture (position of the arms, of

the legs, etc.) of half of the participants in a task inwhich theywere

patron. Firstly, in this context, the behavior evaluated was com-

thisway: Excuseme, Sir/Madam,would you please answer a shortverbal survey in order to improve the quality of the reception of our

C. Jacob et al. / Journal of Retailing and Consumer Services 18 (2011) 381388 383pliance to the sellers commercial suggestion, and this behaviorwasnot altruistic as in previous experiments measuring behavioraldata. Secondly, in this context, it is also possible to measure thecustomers perception of the mimicker and the store, and then tostudy the relationship between the evaluation data and thebehavioral data. Because mimicry was associated with a greaterdesire of afliation, with rapport and a positive perception of themimicker, as well as with persuasion and a higher level ofcompliance toward the mimickers request, we hypothesized thatsales clerks who mimicked customers during their interactionwould be evaluated more favorably by these customers, and thatthe sales clerks suggestions would be more favorably received. Itwas also hypothesized that the store would be more favorablyevaluated in the experimental conditions involving mimicry.

2. Experiment 1

2.1. Method

2.1.1. Setting

The experiment was conducted in a department store offering awide range of products such as domestic electrical appliances, TV,Video, Hi, computers, and CDs. The experiment took place in thesection where MP3-players were sold. In this section, the shelvingwasover8 mlongandmore than80MP3-Playersweredisplayed.Allthe goods on show were for sale. The prices range was comprisedbetween 15h (20$) and 580h (700$). A MP3 player is a productfrequently offered as a gift (birthday, Christmas, etc.), and a lot ofcustomers, particularly people aged 40 or over were requesting thesales clerks help to make a choice. This effect was reinforced by thewide range of products displayed in this section.

2.1.2. Participants and sellers

The participants were 58 males and 74 females (ranging in agefrom 40 to 70), shopping alone, who solicited a sales clerk in theMP3 section for his help in choosing a model. The data of threeparticipants were removed because they refused to respond to thesurvey conducted as they came out of the store. The experimentinvolved four sales clerks (4 young men aged 2228), with a goodasked to evaluate various advertisements. The experimenter, whowas seated in front of the participant, mimicked, or not, theparticipants posture. When the task was concluded, the experi-menter accidentally dropped six pens on the oor. It was foundthat participants in themimicry condition picked up the pensmoreoften (100%) than participants in the non-mimicry condition (33%).Behavioral mimicry can also facilitate the outcome of negotiations.In a recent study byMaddux et al. (2008) it was found thatmimicryfacilitated a negotiators ability to uncover underlying compatibleinterests, and also increased the likelihood of closing a deal in anegotiation where a prima facie solution was not possible.

All above, these studies show that mimicry seems to enhancesocial relationships. For Lakin et al. (2003), the relationshipbetween mimicry and liking or pro-social behavior could beexplained in terms of human evolution. For these authors, mimicrymay serve to foster relationships with others. This behavior couldserve as a social glue, binding people together and creatingharmonious relationships.

1.7. Objective and hypothesizes

The purpose of our experiment was to explore the role ofmimicry in a selling context, with a relationship between a patronand a seller in which the seller was instructed to mimic, or not, asales experience (more than 12 months) with this type of product.patrons. It will only take two minutes of your time. All but three(all three in the non-mimic condition) accepted the survey. Fivenine-point bipolar semantic scales were used to evaluate the salesclerk and the store itself. For the evaluation of the seller, thecustomer was asked to assess the sellers competence, agreeable-ness, ability to stay tuned to the customer, and friendliness. The lastscale was used to make a general evaluation of the store. Theinterviewer thanked the patron, noted his or her sex on the store/seller-evaluation card, and returned into the store. By describingthe customers appearance to the seller, he ascertainedwhether theAs we previously used the same employees to conduct furtherexperiments on the lure technique (Gueguen and Jacob, 2008) orthe tactile contact technique (Gueguen and Jacob, 2008), precau-tions were taken in the later studies to select sales clerks with thesame level of physical attractiveness. A separate sample of 21women and 19 men judged 11 sales clerks on a 1 (not at allattractive) to 9 (extremely attractive) scale. The sales clerks thatseemed to bemoderately attractive both formen andwomenwereselected. Pairwise comparison of the level of attractivenessbetween these four sales clerks revealed non-statistical difference.

2.1.3. Procedure

According to a random distribution, when a customer aged 40 orover solicitedhim forhishelp in choosing aMP3-Player, the sellerwasinstructed to mimic, or not, the customer. The seller began theinteraction by asking the customer to give him the price he/she waswilling to pay. Then the seller suggested a model. In the mimicrycondition, the seller was instructed to mimic the customers verbalbehavior by literally repeating some of his or her words, verbalexpressions or statements. For example, some of the customersinitiated the interaction by saying Hello! Could you help me to chooseaMP3player? In thenon-mimicrycondition, the sellerwas instructedto say Yes of course, whereas in the mimicry condition, he said:Hello! Of course I can help you to choose aMP3 player. If the customersaid: I want to buy a MP3 player for my grandson, then in the non-mimicry condition the seller was instructed to say Yes. How old ishe?, whereas in the mimicry condition, he said AMP3 player for yougrandson. How old is he?. The seller was also instructed to repeat, ornot, some of the customers verbal expressions when presenting aMP3-Player (e. g.: Its funny, Its lighty). In themimicry condition,the sellerwas instructed to try to repeat ve expressions or sentencesduring the interaction. The confederate was also instructed to mimicthe customers nonverbal behavior (i.e. when he stroked his face,foldedhis arms, scratchedhis ears, etc.) during the interaction.Hewasalso instructed to try to mimic nonverbal behavior ve times duringthe session, with a delay of three or four seconds after the customerproduced this nonverbal behavior. In the non-mimicry condition, theseller was instructed to be careful not to mimic verbal expressions orsentences, or the nonverbal behavior of the customer. Outside of thisdifference in verbal andnonverbal behavior in themimicry condition,the four sales clerks were instructed to attempt to act with thecustomers as they usually did: smiling as usual, standing where theyusuallydid in relation to the customer, responding in the sameway. Inboth conditions, the sellers were instructed to present the same twomodels to the customer, according to his/her price range.

A male observer standing alone near the MP3 department wasinstructed to observe the customers discreetly, and to wait untilone of them solicited the seller. Hewas then instructed to approachthe dyad and to ascertain whether the customer was reallysoliciting the seller for information about a MP3 player. Then theobserver left the store and waited until the customer emerged, atwhich point he approached the customer and solicited him/her inpatron had bought a MP3 player, and noted whether the model

selected by the patron was or was not one of the two modelssuggested by the seller.

2.2. Results

In this experiment, 66 customers (29males and37 females)weretested in themimicryexperimental conditionand63 (29menand34women) in the non-mimicry control condition. As no difference inpatron behavior and evaluation was found in the two experimentalconditions, according to the gender of the customer or the sellers,data were collapsed across the sex of patrons and sellers. The rstbehavioral dependent variable used in this experiment was eval-uated by the number of customers who solicited a seller forinformation about a MP3-Player, and who bought a unit. In themimicry condition, 78.8% (52/66) of the sellers solicitors left thestore after buying a MP3-Player whereas they were only 61.8% (39/63) in the non-mimicry condition. The difference between the tworateswas signicant (w2 (1/129)4.42,po0.04, r0.18): customerswhoweremimicked by a seller bought aMP3 playermore favorablythan customers who were not mimicked. The second behavioralvariablemeasured in this experiment considered themodel ofMP3-Player bought by the customers. Indeed, when interacting with thecustomer, the seller was instructed in both conditions to behave ashe always did and to suggest twomodels to the customers.We thenevaluatedwhether the customerwho bought aMP3-player took oneof the twomodels suggestedby the seller, or chose anothermodel. Inthe mimicry condition 71.1% (37/52) of the customers chose one of

the 2 models suggested by the seller, whereas they were only 46.2%(18/39) in the non-mimicry condition. Again, the difference betweenthe two rates was signicant (w2 (1, 91)5.83, po0.02, r0.25). Itthus appears that customers whowere mimicked by a seller boughtmore favorably one of the two MP3 players suggested by the sellerthan customers who were not mimicked.

Once out of the store and in the car-park, the customers weresolicited to evaluate the seller and the store. The data providedfrom the scales are presented in Table 1.

The difference between the two groups was analyzed by thehelp of the Students t-test. Furthermore, when multiple statisticaltests were used with the same data set, we increased the prob-ability of a Type I Error. Accordingly, Bonferronis adjustment wasrecommended to compensate for this increased probability. Onemethodused here consisted in dividing the alpha level (0.05) by thenumber of separate pair-wise comparisons. In this experiment, thealpha level with Bonferronis adjustmentwas equal to 0.01, and thet-value for double-sided testing (df127) was equal to 2.36. In allthe pair-wise comparisons, the calculated t-value was greater than2.36, showing that the differences in evaluation between the twoexperimental conditions were highly signicant. As we can see,mimicrywas associatedwith amore positive evaluation of both the

Table 1Means (SD in brackets) of the evaluation scales according to the experimental

conditions (scores can theoretically vary from 1 to 9 with a higher score associated

with a better positive evaluation).

Mimicry No mimicry Test

Competent 7.70 (0.86) 6.90 (0.82) t(127)5.36, po0.001 d0.99

exp

A

r

p

r

p

r

p

r

p

l co

ed b

C. Jacob et al. / Journal of Retailing and Consumer Services 18 (2011) 381388384a For experimental condition 1Mimicry condition and 2non-mimicry controb For buying behavior 1purchase and 0no purchase.c For choice of the model 1customer had chosen one on the twomodels proposAgreeable 7.62 (0.80) 6.76 (0.76) t(127)6.27, po0.001 d1.11Stay turned to me 7.14 (0.84) 6.44 (0.80) t(127)4.79, po0.001 d0.85Friendly 7.29 (0.72) 6.40 (0.75) t(127)6.88, po0.001 d1.22General evaluation

of the store

7.06 (0.58) 6.35 (0.74) t(127)6.01, po0.001 d1.07

Table 2Correlation matrix between the evaluation-score of the seller and the store and the

Buying behavior Choice of

the model

Competent

Experimental

conditionar(128)0.19po0.04

r(90)0.25po0.02

r(128)0.43po0.001

Buying behaviorb r(128)0.13ns

r(128)0.61po0.001

Choice of the modelc r(90)0.52po0.001

Competent

Agreeable

Stay turned to me

Friendly

General evaluation of

the storeproposed by the seller.sellers and the store.In order to study the relation between these dependent vari-

ables, a correlation analysis (BravaisPearsons coefcient) wasperformed. The correlation matrix is presented in Table 2.

Ascanbeseen, all theevaluationscalesof thesellerwere foundtobehighly and positively correlated, and these personal evaluation scaleswere associated with the general evaluation of the store. The alphacoefcient for the ve itemswas 0.863, suggesting that the items haverelatively high internal consistency. Considering this internal consis-tency, a composite evaluation score was calculated by adding the vescales. It was found in a previous study that this composite score washighly related with customer satisfaction toward the employees andthe store and that this composite score was a good predictor ofcustomer loyalty toward the store (Gueguen and Jacob, 2008).

To test whether the evaluation mediates the effect of mimicry oncustomers behavior, a statistical mediation analysis (MacKinnon,2008)wasperformed.The resultsof themediationanalysis consideringthe customer behavior are illustrated in Fig. 1, whereas the mediationanalysis considering the choice of themodel proposed by the seller areillustrated in Fig. 2.

erimental condition.

greeable Stay turned

to me

Friendly General evaluation

of the store

(128)0.49o0.001

r(128)0.39po0.001

r(128)0.52po0.001

r(128)0.48po0.001

(128)0.51o0.001

r(128)0.47po0.001

r(128)0.39po0.001

r(128)0.43po0.001

(90)0.48o0.001

r(90)0.38po0.001

r(90)0.19po0.001

r(90)0.23po0.04

(128)0.73o0.001

r(128)0.57po0.001

r(128)0.51po0.001

r(128)0.48po0.001

r(128)0.72po0.001

r(128)0.54po0.001

r(128)0.49po0.001

r(128)0.53po0.001

r(128)0.48po0.001r(128)0.53po0.001

ndition.

y the seller and 0customer had chosen another model that one of the twomodels

3. Experiment 2

3.1. Method

3.1.1. Overview

Experiment 2 was similar to Experiment 1 with three excep-tions. Experiment 2was conducted in one other store. However, thestore was near the same store where experiment 1 was carried outand belonged to the same business group. Five sellers participated

B = 0.10, SE = 0.01 = 0.550, t = 4.60, p

C. Jacob et al. / Journal of Retailing and Consumer Services 18 (2011) 3813883864. General discussion

The results of the two experiments presented above support thepostulate that mimicry is associated with customer behavior andcustomer judgment in a real buying context. First, when consideringthe customers behavior, it was found that mimicry was associatedwith a signicant rate of buying behavior, and when this behavioroccurred, it was found that mimicry was associated with a greaterpersuasion effect, because the rate of customers who bought one ofthe MP3 player models suggested by the sellers was signicantlyhigher than when no mimicry was performed. Considering theliterature onmimicry, these results appear to be original. Indeed, inthe only two experiments designed to examine customer behavior,Van Baaren et al. (2003) found that waitresses who mimicked theverbal behavior of customers in a restaurant increased the size of thetips they received. This experiment was conducted in the Nether-lands, where service charges are included in the bill. Thus, giving atip to a waiter or a waitress is not systematic. Giving a tip to arestaurant employee is not strictly speaking an ordinary part ofconsumer behavior, but rather a formof altruism. In our experiment,we evaluated an aspect of consumer behavior which was notassociated with altruism because the customers were interestedin a product, and the dependent variables used in our experimentonly considered whether each customer bought or did not buy theproduct after their interactionwith the seller. The customerbehaviormeasuredhere is clearly original, andgenuinelydifferent than inVanBaaren et al.s study, in which customer behavior seems to be aninappropriate concept: buying behavior appeared to be a moreappropriate form of customer behavior. Thus, for the rst time, wefound thatbuyingbehavior is affectedbymimicry,particularlywhenthis behavior is displayed in a eld setting.We also found that in themimicry condition, customers complied more favorably with theproduct suggestion displayed by the seller. In someways, this resultis congruent with the results found by Bailenson and Yee (2005).These scientists found that a participant interacting with a virtualagent presenting a verbal agreement, found this virtual agent morepersuasivewhenmimicrywas used (the virtual agentmimicked theparticipants head movements at a 4-s delay). It was found thatparticipants indicated a higher agreement rate with the messagedelivered by the agent when they were mimicked. Our experimentbrought out the same persuasive effect on the part of the mimicker,but it was associated with true buying behavior, and not with anagreement with the sellers argument or with buying intent.

In our rst experiment, it was also found that mimicry wasassociated with a more positive evaluation of the mimicker. Suchresults are congruent with previous experimental studies. Chartrandand Bargh (1999, study 2) found that, participants who weremimicked by a confederate subsequently reported a higher meanof liking for the confederate, and described their interaction with theconfederate as smoother and more harmonious compared to thosewhowere notmimicked.Maurer and Tindall (1983) also found that acounselor who mimicked the arm and leg position of a client wasevaluated more favorably by the client. Our results conrm theseprevious data by showing that mimicry enhanced liking for themimicker in a selling context. Perhaps this effect could be explainedby the fact that mimicry as a process establishes a higher-qualityrapport: a rapport which, in return, affects the evaluation of themimicker and his/her powers of persuasion. LaFrance (1979) foundthroughcorrelational analysis that a synchronyofposturebetweenaninstructor and his/her colleges studentswas associatedwith a higherevaluation score of the class rapport with the instructor. Lakin andChartrand (2003) found that mimicry may be part of a behaviorrepertoire designed to create rapport with counterparts. Theseauthors demonstrated that having an afliation goal, or havingunsuccessfully attempted to afliate in an interaction, increased

mimicry. In the same way, Johnston (2002) and Yabar et al. (2006)found greater mimicry of amember of an in-group than of amemberof an out-group. If this is the case, and the desire of afliation andmimicryare so clearlyassociated, this relationcouldhelpus toexplainour results. Ifwemimic others in order to create afliation,wemaybeable to perceive that others who mimic us are expressing a desire tocreate afliation. Thus, in our experiment, mimicry could haveenhanced the customers feeling of a greater rapport with the seller.One of the evaluation scales used in this experiment (did not stay/stayed tuned to me) seems to conrm that mimicry enhanced thisperception of the sellers desire to create afliation and rapport withthe customer. An additional effect of mimicry on judgment wasobtained in our experiment when it was found that the evaluation ofthe mimickers professional competence was positively affected bymimicry. This effect is new, because this dimension has never beenevaluated in previous studies. This result could not be explained by ahalo effect, because it was the rst scale presented verbally to thecustomer by the interviewer. Such a result is congruent with thetheoretical position of Lakin et al. (2003), forwhom the positive effectofmimicry in social interaction is explainedby the fact thatmimicry isinterpreted by the counterpart as a desire, on the mimickers part, tocreate afliation and rapport. Now, in the part of our experiment inwhich customers had to judge a seller, this desire to create afliationand rapport could have been perceived as professional competence,because this desire is perceived as an important quality for a seller toposses if he wishes to improve his sales performance. With somenonverbal behavior like tactile contact, it has been found that a salesclerk or restaurant employee giving a slight touch to a customer, isassociated with a higher buying rate, a greater amount of moneyspent, and larger tips given to anemployee (Gueguenand Jacob, 2006;Hornik, 1992a; Kaufman and Mahoney, 1999; Smith et al., 1982).Concomitantly, the professional qualities of the seller or the restau-rant employee were evaluated more positively (Crusco and Wetzel,1984; Hornik, 1992b; Stephen and Zweigenhaft, 1986; Wycoff andHolley, 1990). As in our experiment, it was also found that asuggestion made by a waiter was more favorably accepted whenthe employee slightly touched the patrons when interacting for therst time (Gueguen et al., 2007). In these experiments, the variouseffects of tactile contact were interpreted as a consequence of theemployees desire to create a better rapport with the customer. Inreturn, thisperceptionby thecustomerwill explainwhy thecustomerbought more products or gave more tips, and why the toucher wasevaluated more positively. If mimicry, like tactile contact, has theability to create this impression of a greater desire on the part of themimicker to create afliation and rapport with the personmimicked,this could explain our results on behavior and judgment.

Another interesting effect found in this experiment was thepositive effect of mimicry on the evaluation of the store. Previousstudies never evaluated this perception of the context. Of coursethis effect is perhaps explained by a classical halo effect, but it isalso possible that mimicry has the potential to create a positivemood, leading in turn to a more positive evaluation of the store. Arecent study found thatmimicrywas associatedwith an increase inpositive mood. Gueguen (2009) carried out an experiment in twobars during sessions of speed dating for which young womenconfederates volunteered to mimic or not some verbal expressionsand nonverbal behaviors of a man for 5 min. Data revealed that themen evaluated the dating interaction more positively when thewoman mimicked them, and that mimicry was associated with ahigher evaluation score of the relation. It was also found thatmimicry increased mens mood. To return to touch, its effect onmood has been shown in previous studies (Fisher et al., 1976;Whitcher and Fisher, 1979). Thus, given the potential ofmimicry tocreate the same feelings and to affect customer behavior in thesameway, it will be interesting to explore this effect of mimicry onmood in future studies. Tactile contact by someone interactingwith

us has the property to inuence our mood and is perceived as a

of mimicry on behavior was tested (Van Baaren et al., 2003, 2004)

Gwinner, K.P., Gremler, D.D., Bitner, M.J., 1998. Relational benets in servicesindustries: the customers perspective. Journal of the Academy of Marketing

C. Jacob et al. / Journal of Retailing and Consumer Services 18 (2011) 381388 387this mediation effect of judgment was not tested. Further studiesalso found that mimicry was associated with more favorablejudgement of the mimicker (Chartrand and Bargh, 1999; Maurerand Tindall, 1983) but without introducing any behavioral mea-sure. In this experiment, for the rst time, this mediating effect ofjudgement in the relation between mimicry and behavior wastested and tested in a real-life situation. Thus, our results seems toshow thatmimicry led to increase positive feelings of themimicker,that led the person mimicked, in return, to comply with themimickers suggestion.

4.1. Limitations

The present study has certain limitations that need to be takeninto account when considering the study and its contributions. Thesample sizes were low and a replicationwith larger samples is nownecessary to generalize the ndings. Only customers aged 40 orover were tested in these experiments for practical reason (if wasfound that in this store, this age range often requested informationand help from the seller) but the generalization of the effect ofmimicry to customers below 40 years is necessary. In this experi-ment we tested only four clerks. This number is larger than in VanBaaren et al. (2003) who used only one employee but replicationwith a larger sample of employees is necessary and with a bettercontrol of some variables associated with the employees: age,gender, physical attractiveness. Moreover, generalization of theeffect of mimicry to different products and different stores is nowessential. We found a positive effect of mimicry on customerbehavior and judgment in these experiments conducted in France,but the generalization of such a positive effect to other cultures stillremains in question given the fact thatmost of the previous studieson mimicry were conducted in western countries. At least, as wefound that mimicry is associated with positive evaluations of thesellers and the store, it would be interesting for further studies toevaluate the effect of mimicry on other aspects of customerevaluations, such as perception of service quality, customer truth.In these experiments, we only studied the immediate effect ofmimicry on customer behavior, but it would also be interesting tostudy the long-term effect, such as loyalty or word-of-mouthbehavior.

4.2. Managerial applications

Our study would appear to present clear managerial applica-tions. It might be nancially advantageous for store managers toincite their sales staff to use mimicry when interacting withcustomers, in order to create in them this greater desire ofafliation and rapport, but also to improve the evaluation of thestore and to improve sales. Such positive effects on evaluationcould be interesting in the creation of a positive word of mouth(East et al., 2008) and/or to reinforce customer loyalty (Gremler andBrown, 1996).

5. Conclusion

In conclusion, for the rst time in an experimental approach, itwas found in this experiment that mimicry was associated with agreater behavioral inuence in a selling context. This effect ondesire to create afliation and rapport. The same effect may beactivated by mimicry.

More interesting, it was found that customers judgement of thestore and the seller mediated the relation between mimicry andbuying behavior. To our knowledge, in the studies where the effectcustomer behavior observed in a real selling context was theScience 26, 101114.Harakeh, Z., Engels, R., van Baaren, R.B., Scholte, R.H.J., 2007. Imitation of cigarette

smoking: an experimental study of smoking in a naturalistic setting. Drug andAlcohol Dependence 86, 199206.

Hornik, J., 1992a. Effects of physical contact on customers shopping time andbehavior. Marketing Letters 3, 4955.

Hornik, J., 1992b. Tactile stimulation and consumer response. Journal of ConsumerResearch 19, 49458.

Johnston, L., 2002. Behavioral mimicry and stigmatization. Social Cognition 20,1835.

Kaufman, D., Mahoney, J.M., 1999. The effect of waitresses touch on alcoholconsumption in dyads. The Journal of Social Psychology 139, 261267.

LaFrance, M., 1982. Posture mirroring and rapport. In: Davis, M. (Ed.), Interactionrhythms: Periodicity in Commutative Behavior. Human Sciences Press,rst evaluation of this type of the effect of mimicry. These resultsalso conrmed previous experimental studies showing that mimi-cry was associated with greater score of liking for the mimicker.Furthermore, these studies focused on the assessment of personalqualities. In our experiment, it was found that the assessment ofother qualities was signicantly improved by the use of mimicry,such as a higher score for the sellers professional qualities, and ahigher evaluation of the context. This effect tends to show thatmimicry may be a powerful new inuence technique, of interest toboth scientists and business and sales professionals.

References

Bailenson, J.N., Yee,N., 2005. Digital chameleons: automatic assimilationofnonverbalgestures in immersive virtual environments. Psychological Science 16, 814819.

Bernieri, F.J., 1988. Coordinated movement and rapport in teacherstudent inter-actions. Journal of Nonverbal Behavior 12, 120138.

Bitner, M.J., 1990. Evaluating service encounters: the effects of physical surround-ings and employee responses. Journal of Marketing 54, 6982.

Bush, L.K., Barr, C.L., McHugo, G.J., Lanceta, J.T., 1989. The effects of facial control andfacial mimicry on subjective reaction to comedy routines. Motivation andEmotion 12, 120138.

Cappella, J.N., Panalp, S., 1981. Talk and silence sequences in informal conversations:III. Interspeaker inuence. Human Communication Research 7, 117132.

Chartrand, T.L., Bargh, J.A., 1999. The chameleon effect: the perception-behavior linkand social interaction. Journal of Personality and Social Psychology 76, 893910.

Cialdini, R., 2011. Inuence: Science and Practice. Allyn & Bacon, Boston.Coulter, K.S., Coulter, R.H., 2000. The effects of service representative characteristics

on trust: the moderating role of length of relationship. In: Gundlach, C.T.,Murphy, P.E. (Eds.), AMA Summer Educators Conference Proceedings: Enhan-cing Knowledge Developement inMarketing. AmericanMarketing Association,Chicago, IL, pp. 12.

Crusco, A.H., Wetzel, C.G., 1984. The Midas touch: the effects of interpersonal touchon restaurant tipping. Personality and Social Psychology Bulletin 10, 512517.

East, R., Hammond, K., Lomax, W., 2008. Measuring the impact of positive andnegative word of mouth on brand purchase probability. International Journal ofResearch in Marketing 25, 215224.

Ebesu Hubbard, A.S., Tsuji, A., Williams, C., Seatriz, V., 2003. Effects of touch ongratuities received in same-gender and cross-gender dyads. Journal of AppliedSocial Psychology 33, 24272438.

Erceau, D., Gueguen, N., 2007. The effect of touch on the evaluation of the toucher.The Journal of Social Psychology 147, 441444.

Fisher, J.D., Rytting, M., Heslin, R., 1976. Hands touching hands: affective andevaluative effects on interpersonal touch. Sociometry 39, 416421.

Giles, H., Powesland, P.F., 1975. Speech Style and Social Evaluation. Academic Press,London.

Gremler, D.D., Brown, S.W., 1996. Service loyalty: its nature, importance, and implica-tions. In: Edvardsson, B., Brown, S.W., Johnston, R. (Eds.), Advancing Service Quality:A Global Perspective. International Service Quality Association, pp. 171180.

Gremler, D.D., Gwinner, K.P., Brown, S.W., 2001. Generating positiveword-of-mouthcommunication through customeremployee relationships. International Jour-nal of Service Industry Management 12, 4459.

Gueguen, N., 2009.Mimicry andwomen attractiveness: an evaluation in a courtshipcontext. Social Inuence 4, 249255.

Gueguen, N., Jacob, C., 2005. The effect of touch on tipping: an evaluation in a Frenchbar. International Journal of Hospitality Management 24, 295299.

Gueguen, N., Jacob, C., 2006. Touch and consumer behavior: a new experimentalevidence in a eld setting. International Journal of Management 23, 2433.

Gueguen, N., Jacob, C., 2008. La technique du leurre en situation dachat: impact durenforcement dune decision sur le renoncement. La Revue des Sciences deGestiondirection et gestion des entreprises 234, 115120.

Gueguen, N., Jacob, C., Boulbry, G., 2007. Servers suggestion and tactile contact:an evaluation in a restaurant. International Journal of Hospitality Management26, 10191023.New York, pp. 279298.

LaFrance, M., 1979. Nonverbal synchrony and rapport: analysis by the cross-lagpanel technique. Social Psychology Quaterly 42, 6670.

Lakin, J.L., Chartrand, T.L., 2003. Using nonconscious behavioral mimicry to createafliation and rapport. Psychological Science 14, 334339.

Lakin, J.L., Jefferis, V.E., Cheng, C.M., Chartrand, T.L., 2003. The chameleon effect associal glue: evidence for the evolutionary signicance of nonconsciousmimicry.Journal of Nonverbal Behavior 27, 145162.

Lynn, M., Le, J.M., Sherwyn, D., 1998. Reach out and touch your customers. CornellHotel and Restaurant Administration Quarterly 39, 6065.

Maddux, W.W., Mullen, E.E., Galinsky, A.D., 2008. Chameleons bake bigger pies andtake bigger pieces: strategic behavioral mimicry facilitates negotiation out-comes. Journal of Experimental Social Psychology 44, 461468.

Maurer, R.E., Tindall, J.H., 1983. Effects of postural congruence on clients perceptionof counselor empathy. Journal of Counseling Psychology 30, 158163.

MacKinnon, D.P., 2008. Introduction to Statistical Mediation Analysis. LawrenceErlbaum Associates, New York.

Provine, R.R., 1992. Faces as releasers of contagious yawning: an approach to facedetection using normal human subjects. Bulletin of the Psychonomic Society27, 211214.

Quigley, B.M., Collins, R.L., 1999. Modeling of alcohol consumption: a meta-analyticreview. Journal of Studies on Alcohol 60, 9098.

Smith, D.E., Gier, J.A., Willis, F.N., 1982. Interpersonal touch and compliance with amarketing request. Basic and Applied Social Psychology 3, 3538.

Stephen, R., Zweigenhaft, R.L., 1986. The effect on tipping of a waitress touching

male and female customers. The Journal of Social Psychology 126, 141142.Van Baaren, R.B., Holland, R.W., Kawakami, K., Van Knippenberg, A., 2004. Mimicry

and prosocial behaviour. Psychological Science 14, 7174.VanBaaren, R.B., Holland, R.W., Steenaert, B., VanKnippenberg, A., 2003.Mimicry for

money: behavioral consequences of imitation. Journal of Experimental SocialPsychology 39, 393398.

Webb, J.T., 1972. Interview synchrony: an investigation of two speech ratemeasures. In: Siegman, A.W., Pope, B. (Eds.), Studies in Dyadic Communication.Pergamon Press, New York, pp. 115133.

Whitcher, S.J., Fisher, J.D., 1979.Multidimensional reaction to therapeutic touch in ahospital setting. Journal of Personality and Social Psychology 37, 8796.

Wycoff, E.B., Holley, J.D., 1990. Effects of ight attendants touch upon airlinepassengers perceptions of the attendant and the airline. Perceptual and Motor

Skills 71, 932934.Yabar, Y., Johnston, L., Miles, L., Peace, V., 2006. Implicit behavioral mimicry:

investigating the impact of group membership. Journal of Nonverbal Behavior30, 97113.

C. Jacob et al. / Journal of Retailing and Consumer Services 18 (2011) 381388388

Retail salespeople's mimicry of customers: Effects on consumer behaviorIntroductionInterpersonal bonds in employee-customer relationshipsMimicry in social relationshipsMimicry: an automatic behaviorMimicry increases positive evaluation of the mimickerMimicking someone to create affiliation and rapportBehavioral effects of mimicryObjective and hypothesizes

Experiment 1MethodSettingParticipants and sellersProcedure

ResultsDiscussion

Experiment 2MethodOverview

Results

General discussionLimitationsManagerial applications

ConclusionReferences