You are on page 1of 8

Journal of Retailing and Consumer Services 21 (2014) 897–904

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services

journal homepage:

Turning dissatisfied into satisfied customers: How referral reward

programs affect the referrer's attitude and loyalty toward the
recommended service provider
Madlen Kuester n, Martin Benkenstein
Institute of Marketing and Services Research, University of Rostock, Ulmenstr. 69, 18057 Rostock, Germany

art ic l e i nf o a b s t r a c t

Article history: Referral reward programs are becoming a popular tool for acquiring new customers and bonding
Received 18 March 2014 existing ones. Yet their benefits are contentious, since such campaigns are prone to the opportunistic
Received in revised form behaviour of customers who merely want to reap the reward. This paper examines how participating in a
22 July 2014
referral campaign affects opportunistic recommenders. By conducting two experimental studies, this
Accepted 23 July 2014
Available online 26 August 2014
article shows that giving counterattitudinal referrals enhances the communicator's attitude and loyalty
toward the recommended provider. However, the positive effect depends on the reward size. While
Keywords: referral reward programs with small incentives strengthen the recommender's attitude and loyalty, no
Referral impact was found for referrals with large rewards. The results show that a stronger focus on reward
programs is worth considering, since service providers can benefit from opportunistic customers with
regards to the bonding effect.
Dissonance & 2014 Elsevier Ltd. All rights reserved.

1. Introduction referrals can be driven by motives of self-interest (Wirtz et al.,

2013). In other words, such programs are prone to the opportu-
Buzz marketing has become an important marketing technique nistic behaviour of consumers (Helm, 2000; Schmitt et al., 2011;
to stimulate positive word-of-mouth (WOM), and thus to acquire Wirtz and Chew, 2002) who merely want to take advantage of the
new customers (Rosen, 2001; Tuk et al., 2009). A recent Web provider in order to reap the provided reward (Blazevic et al.,
search using the keyword “member-get-member-campaign” 2013; Wirtz and Chew, 2002). Making stimulated referrals by
yielded 40,000 results, showing that referral programs are gaining unsatisfied customers is not a rare phenomenon (Helm, 2000).
growing popularity in various service industries (Ryu and Feick, Experimental studies confirmed that dissatisfied customers
2007; Schmitt et al., 2011; Verlegh et al., 2013). engage in incentivized referrals only because of the reward
In referral programs, existing customers are rewarded for (Garnefeld et al., 2009). Although scholars agree that WOM
bringing in new customers (Biyalogorsky et al., 2001; Ryu and opportunism exists, very little research has been done on self-
Feick, 2007; Schmitt et al., 2011). In this way, incentives basically interested WOM behaviour. Schmitt et al. (2011) found that the
function as an extrinsic motivator to generate WOM (Tuk et al., contribution margin, the retention rate, as well as the customer
2009; Wirtz and Chew, 2002). Although previous studies con- lifetime value for customers acquired through a referral campaign
firmed that providing rewards increases referral likelihood (Ryu are higher than for non-referred customers, and that the value
and Feick, 2007; Wirtz and Chew, 2002) and the favourability of differential is larger than the reward. These results indicate that
the WOM generated (Wirtz and Chew, 2002), referral reward any misuse by opportunistic consumers is much less important
programs are not uncontroversial (Schmitt et al., 2011; Tuk et al., than the benefits stemming from newly acquired customers
2009). Some academics have questioned the value of incentive (Schmitt et al., 2011). Furthermore, a few studies examined how
referral campaigns (e.g., Trusov et al., 2009; Van den Bulte, 2010; a reward affects receivers' responses to the recommendation.
see also Schmitt et al., 2011), particularly because rewarded Verlegh et al. (2013) showed that incentives adversely affect
responses when they are perceived as driven by ulterior motives.
Tuk et al. (2009) generally speculate that the reward introduces a
Corresponding author. Tel.: þ 49 176 2444 9978; fax: þ49 381 498 4378.
financial motive into the interpersonal interaction between the
E-mail addresses: (M. Kuester), referring and the receiving customer, which could harm the (M. Benkenstein). sincerity of the recommendation. However, researchers have
0969-6989/& 2014 Elsevier Ltd. All rights reserved.
898 M. Kuester, M. Benkenstein / Journal of Retailing and Consumer Services 21 (2014) 897–904

neglected analysing the impact of giving counterattitudinal refer- the evaluation of a service provider (Davidow, 2003; Garnefeld
rals on the referring customer. Recent research into the attitude- et al., 2013; Nyer and Gopinath, 2005). Participating in a referral
congruent WOM context has shown that participating in a referral reward program thus constitutes a public commitment that should
program increases the communicator's attitudinal and behavioural have consistency effects (Blazevic et al., 2013; Garnefeld et al.,
loyalty to the provider (Garnefeld et al., 2013). By contrast, the 2013). Pursuant to Cialdini (2009), dissonance theory and self-
effect of articulating positive WOM that does not reflect the perception theory are eligible theoretical frameworks for explain-
referrer's attitude toward the provider has received no research ing these consistency effects. As shown by Garnefeld et al. (2011),
attention thus far. This study therefore aims to bridge this gap in self-perception theory is suitable for attitude-congruent WOM
the literature by examining the effects of articulating opportunistic behaviour, whereas dissonance theory provides an accurate
referrals on the recommender's attitude and loyalty. In this regard, approach for the effects of attitude-incongruent commitments.
this paper responds specifically to Garnefeld et al.'s (2013) call for We draw on Festinger's (1957) dissonance theory to derive our
research into further investigations of the bonding effect of hypothesis because we focus on the impact of attitude-discrepant
attitude-discrepant WOM behaviours on the referring customer. referrals on the recommender's attitude and behaviour.
The lack of research into the sender-related impact of recom- Pursuant to dissonance theory, individuals strive to implement
mendations is an obvious shortcoming in the WOM literature (Ryu their cognitive system in a state of psychological consistency (Frey
and Feick, 2007; Tax et al., 1993). Only a few studies have explored and Gaska, 2001). Cognitions are elements of knowledge that
the effects of referrals on the recommender (Eggert et al., 2007; individuals have about their attitudes, beliefs, values, and beha-
Garnefeld et al., 2011, 2013). By showing that positive referrals are viour (Cooper and Carlsmith, 2001; Olson and Stone, 2005).
not only effective for gaining but also for bonding customers, these Festinger (1957) asserts that an inconsistency between two rele-
studies added a new perspective on the consequences of WOM vant cognitions generates a psychological discomfort, which he
communication (Garnefeld et al., 2011). Thus, referral reward conceptualized as an aversive state similar to hunger or frustra-
programs can also be regarded as an instrument for retaining tion. That unpleasant tension that Festinger termed “dissonance”
existing customers (Jin and Huang, 2013; Reinartz et al., 2004; Ryu would motivate people to reduce this inconsistency by changing
and Feick, 2007). We broaden this view by investigating the effects one cognitive element (Cooper and Carlsmith, 2001; Olson and
of opportunistic referrals on the recommender's attitude and Stone, 2005).
loyalty. We further examine their boundary conditions; more Several early studies confirmed that engaging in counterattitu-
specifically, the size of the reward is investigated as a moderator. dinal behaviour evokes cognitive dissonance and thus a state of
Owing to the growing use of referral reward programs, a better aversive arousal (e.g., Elkin and Leippe, 1986; Fazio et al., 1977;
understanding of their impact on customer retention is of great Kiesler and Pallak, 1976; Losch and Cacioppo, 1990; Zanna and
interest for service marketing management, both for scholars and Cooper, 1974; Zanna et al., 1976). By giving attitude-incongruent
practitioners. By rewarding WOM, service providers invest both in referrals, customers should experience a psychological discomfort
acquiring and in retaining satisfied customers (Ryu and Feick, because the behavioural cognition – “I recommended my service
2007). However, dissatisfied customers who are driven by the provider” – and the attitudinal cognition – “I don't like my service
reward are also attracted. Referral campaigns are very cost provider” – are dissonant. Due to this public advocating, the
effective (Schmitt et al., 2011; Xiao et al., 2011), and the spending recommender has committed him- or herself to his or her
on those programs has increased considerably (Blazevic et al., behaviour. Brehm and Cohen (1962) postulates that a commitment
2013). Therefore, the uncertainty that surrounds the payback of to a dissonant cognition makes it resistant to change, and the
customer referral programs due to opportunistic participants is recommender is thus tied to his or her referral. Owing to this
frustrating for managers aiming to prevent wasteful expenses and commitment to the counterattitudinal behaviour, there should be
to boost their return on marketing investment (Schmitt et al., no possibility for the participant of the referral program to reverse
2011). For a deeper insight into the effectiveness of customer or renege on his or her behaviour, or to distort its meaning.
referral programs, companies need to understand how the parti- Attitude change should therefore remain the only viable option for
cipation in a campaign affects opportunistic customers, and eliminating dissonance (Eagly and Chaiken, 1993). Accordingly, it
whether providers might even benefit from a bonding effect. By is expected that referrers would change their attitude toward the
exploring the effects of campaigns with both small and large recommended service provider in order to reduce the inconsis-
rewards, we provide further guidance on how to design referral tency between attitude and behaviour. From this discussion, the
reward programs. following hypothesis is advanced:

H1. Articulating a counterattitudinal referral within a referral

2. Theoretical background and research hypotheses reward program has a positive effect on the recommender's
attitude toward the service provider.
2.1. Communicator's attitude
2.2. Communicator's loyalty
Individuals who advocate a specific position are driven to be
consistent with their articulated position (Cialdini, 1971, 2009; A public disclosure of an attitudinal position strongly influ-
Cialdini and Goldstein, 2004; Garnefeld et al., 2013). That binding ences subsequent behaviour (Cialdini, 2009); in other words,
to a position results from the public commitment implied by the individuals are more likely to choose actions congruent with one's
subjects' behaviour (Kiesler, 1971; Nyer and Dellande, 2010). Public public commitments (Cialdini and Goldstein, 2004). Eagly and
commitment occurs when an individual's position is known or Chaiken (1995) states that commitment drives people into a state
made public to other people (Gopinath and Nyer, 2009). Research of defence motivation. Defence motivation leads to selective
provides empirical evidence that persons who make public com- cognitive processing of information that is inconsistent with one's
mitments align their attitudes and future behaviours with their public commitments (Gopinath and Nyer, 2009; Pomerantz et al.,
commitments (e.g., Cioffi and Garner, 1996; Garnefeld et al., 2011; 1995). According to dissonance theory, people desire to avoid
Greenwald et al., 1987; Katzev and Pardini, 1987; Nyer and increases in dissonance by evading dissonant information (Frey
Dellande, 2010; Pallak and Cummings, 1976; Wang and Katzev, and Gaska, 2001; Olson and Stone, 2005). Individuals are therefore
1990). By generating WOM, the customer takes a public stand on more resistant to the communication efforts of other service
M. Kuester, M. Benkenstein / Journal of Retailing and Consumer Services 21 (2014) 897–904 899

providers (Gopinath and Nyer, 2009), and thus behave consistent customers to receive various benefits to maintain and restore their
with their commitments (Cialdini and Goldstein, 2004). If a health was chosen as the research context because referral reward
customer recommends a service provider but then switches firm, programs are well-established in the insurance industry. Further,
the customer would perceive that action as inconsistent with his participants should be familiar with statutory health insurance, as
or her public commitment (Garnefeld et al., 2013) and experience they are basically obligated to take out such insurance.
a psychological discomfort. Participating in a referral reward
program should therefore increase the intention to stay with the
3.2. Study 1
recommended service provider. Hence, we hypothesize:

H2. Articulating a counterattitudinal referral within a referral 3.2.1. Research design and participants
reward program has a positive effect on the recommender's We experimentally created two groups: a treatment group
loyalty toward the service provider. which articulated a counterattitudinal referral, and a control group
which did not. Before we manipulated giving a counterattitudinal
referral, we manipulated a negative attitude toward a service
2.3. The moderating effect of reward size
provider by exposing both the treatment and the control group to
a situation in which they have a negative business relationship with
Festinger's (1957) theory predicts that the magnitude of dis-
a health insurance provider (attitudinal scenario). To manipulate
sonance depends on the incentive used to induce a person to
articulating a recommendation, we followed Eggert et al.'s (2007)
advocate a counterattitudinal position (Eagly and Chaiken, 1993).
approach in the attitude-congruent WOM context, and presented
According to dissonance theory, the greater the inducement
the treatment group with a further description of a situation in
provided for an attitude-discrepant behaviour, the less the resul-
which they recommend the service provider (referral scenario).
tant dissonance (Carlsmith et al., 1966; Stone and Fernandez,
We pretested the attitudinal scenario with 18 students to
2008). Having participated in a referral program with a large
ascertain the effectiveness of the manipulation of a negative
reward, the size of the incentive would provide a sufficient
attitude. Pretest respondents were further questioned on the
justification for the attitude-incongruent referral, and thus add a
referral scenario. In particular, they were asked to summarize
significant consonant cognition to the element representing the
the scenario description, and to comment if they had problems
counterattitudinal behaviour (Cooper and Carlsmith, 2001; Eagly
understanding the simulated situation. Pretest feedback showed
and Chaiken, 1993). As a result, the recommender should experi-
that participants were able to project themselves well into the
ence only little dissonance in large-rewarded campaigns. There-
described situations.
fore, customers would not need to change their attitude and their
For data collection, a total of 110 questionnaires were distrib-
loyalty toward the service provider in the direction of the
uted to students at a northern European University. Respondents
articulated position. By contrast, referring customers who receive
did not receive any compensation for their participation in the
a small reward for participating in a referral reward program
study. Students were randomly assigned to one of the two groups.
should perceive this small incentive as an insufficient rationale for
11 questionnaires were excluded due to incomplete data, resulting
their counterattitudinal advocacy, and experience a psychological
in 99 acceptable response sets for analysis.
discomfort due to the articulation of a counterattitudinal referral.
Hence, in the case of small-rewarded referrals, communicators
would become motivated to reduce the dissonance arousal by 3.2.2. Experimental procedure
increasing their attitude and loyalty. Based on these considera- Firstly, all participants received a written scenario comprising a
tions, we derive the following hypothesis: description of the subject's business relationship with the fictitious
statutory health insurance company “Health-Plus” (see Fig. A1). In
H3. The positive effects of articulating a counterattitudinal referral the scenario, the relationship was described in negative terms in
within a referral reward program on the recommender's attitude order to induce an adverse attitude toward the provider. That
and loyalty toward the service provider are moderated by the size negative attitude was created by references to the unpleasant
of the reward: the positive effects only arise in a referral reward bilateral contact with the provider, and additional fees which have
program with small rewards. to be paid for the insurance coverage. Students were instructed to
read the description carefully, and try to project themselves into the
3. Methodology hypothetical situation. Subsequently, the participants rated their
attitudes toward the service provider on a bipolar three-item scale
3.1. Overview ranging from 1 (“very negative, very unfavourable, very bad”) to 7
(“very positive, very favourable, very good”) (Laczniak et al., 2001;
We conducted two experimental studies with a between- Simmons and Becker-Olsen, 2006). This was so that we could test
subjects design to test our hypotheses. Study 1 investigated the the manipulation of the negative attitude, and also determine
effects of articulating a counterattitudinal referral on the commu- whether both groups produced similar attitudes. Before the students
nicator's attitude and loyalty toward the service provider (H1 and received a second questionnaire packet containing the referral
H2). In a follow-up study, we replicated the effects from the first scenario and several rating questions, we collected the attitudinal
experiment, and further tested the moderating impact of the reward scenario, including students' responses, as we needed to examine
size (H3). In our experiments, we used a scenario approach to the impact on the communicator's attitude. We thus avoided
manipulate a counterattitudinal referral (see Fig. A1 in the appendix participants just copying the pretest scores from the first scenario.
for a scenario description). Presenting participants with a descrip- In the second questionnaire packet, only the treatment group
tion of a fictitious situation and asking them to picture themselves in received a scenario that described them giving a referral, while the
that situation is a widely used technique in the marketing field control group was just asked to answer the questions. In the
(Wirtz and Chew, 2002). In WOM research, this method has been referral scenario, students were asked to imagine meeting a friend
successfully applied in a number of studies (e.g., Garnefeld et al., at a coffee bar, who mentions a desire for a new health insurance
2011, 2013; Laczniak et al., 2001; Söderlund and Rosengren, 2007; provider (see Fig. A1). We also asked them to imagine that they
Wang, 2011; Wirtz and Chew, 2002). A statutory health insurance had recently received a flyer about a member-get-member-
service which provides conventional coverage and entitles campaign from their insurance company, “Health-Plus”, with the
900 M. Kuester, M. Benkenstein / Journal of Retailing and Consumer Services 21 (2014) 897–904

request to recommend the provider to friends. The participants 3.3. Study 2

were informed that they would receive a monetary reward for
every new customer they successfully acquired. In order to 3.3.1. Research design and participants
increase the realism and the imaginativeness of the scenario, the In the follow-up experiment, we extended the experimental
description was accompanied by a picture depicting two persons design from Study 1 to two different levels of reward size (small
sitting in a café. To manipulate participation in a referral reward and large). The referral scenario of the first study was slightly
program, the students of the treatment group were then asked to adapted (see Fig. A1). An amount of €10 was chosen as a small
articulate their referrals in a speech bubble. The picture reward, and €100 as a large reward. These reward sizes are
(Söderlund and Rosengren, 2007), as well as the text box common in referral campaigns of European health insurance
(Garnefeld et al., 2011, 2013), should enhance the vividness and providers, and present the actual range of incentives offered for
tangibility of the described situations. participation in health insurance referral campaigns. We pretested
Both groups responded to some questions regarding the different reward sizes to ascertain whether the sizes are perceived
dependent variables using established scales adapted to our as realistic and that they correspond with the assumption of
research context. For measuring participant's attitudes, we used dissonance theory. The selected amounts emerged as meeting
the same seven-point scale as within the attitudinal scenario. the requirements.
Based on the study of Ganesh et al. (2000), we measured loyalty We thus created three experimental conditions in Study 2.
using four items (on a scale of 1 ¼“strongly disagree” to While the control group was only exposed to the attitudinal
7 ¼“strongly agree”). Finally, all participants were asked to rate scenario, the participants assigned to the treatment conditions
how well they could project themselves into the situation on a also received a description about recommending their service
seven-point scale (1 ¼“the situation was very easy to picture” to provider for either a small or a large reward. Study participants
7 ¼“the situation was very difficult to picture”). A complete list of were recruited from business courses at the same northern
the items used to measure the subjects’ attitude and loyalty is European University as in Study 1. A total of 120 students took
provided in Table A1 in the Appendix. The scales for measuring the part in the experiment, and were randomly allocated to one of the
participants' attitude and loyalty performed well on the tests three experimental conditions. The sample distribution across the
conducted to assess the quality of the instruments (see Table A1). three experimental cells ranged from 39 to 41 respondents.
Students did not receive any compensation for their participation.
3.2.3. Manipulation and realism checks
We confirmed that the attitudinal scenario induced a negative 3.3.2. Experimental procedure
attitude toward “Health-Plus” (M(99) ¼1.77). As intended, no As with the procedure in Study 1, each participant first read the
significant differences were found between the control and the scenario for manipulating a negative attitude, and then rated their
treatment group (t(99) ¼  .91, p ¼.37, MCG ¼1.69, MTG ¼1.84). The attitudes toward the service provider “Health-Plus”. After collect-
realism check was also successful and showed a mean of 3.60. ing the scenario and the response sets, the students received the
Thus, participants could imagine experiencing the presented second questionnaire packet comprising the referral scenarios, as
situation. Both groups perceived the realism of the scenarios well as numerous questions relating to the dependent variables
similarly (t(99) ¼  .37, p¼ .71, MCG ¼3.54, MTG ¼ 3.65). Further, we and manipulation checks. Participants in the small reward group
ensured that participants in the referral group filled out the speech were informed that they would receive a reward of €10 if they
bubble and gave positive referrals. All students articulated their successfully recommended the provider. In the large reward
referral by writing such statements such as “I can recommend you group, subjects read that “Health-Plus” would reward a referral
my insurance provider” or “Switch to my firm. If you sign up, I with €100. As in Study 1, the participants assigned to the control
would get a reward. I would invite you for coffee next time”. group did not receive the referral scenario.
Again, all groups completed a questionnaire. In Study 2, we
used the same construct measures for attitude and loyalty as in
3.2.4. Findings
Study 1. Besides the realism check from Study 1, we also included a
Due to the correlation among the dependent variables (r ¼.51,
check for the manipulation of reward size and the magnitude of
p o.001), we conducted a multivariate analysis of covariance
dissonance. To test the reward size manipulation, we used two
(MANCOVA) to test the hypothesized effects of articulating a
seven-point semantic differential items developed by Ryu and
counterattitudinal referral on the recommender's attitude (H1)
and loyalty (H2). The pretest score of participants' attitudes Feick (2007): 1 ¼“very small amount, very unattractive amount”
and 7 ¼ “very high amount, very attractive amount”. For checking
toward the service provider was included as a covariate in the
model. Table 1 presents the results of the analysis. In support of the magnitude of dissonance, we asked participants from the
reward conditions whether the articulation of the referral made
H1, we confirmed a significant positive effect on the communica-
tor's attitude (F(1,99)¼ 21.82, p o.001, partial η2 ¼.19, MCG ¼1.95, them feel uncomfortable (on a scale 1 ¼“very uncomfortable” to
7¼ “not uncomfortable at all”; Fazio et al., 1977). The tests for
MTG ¼ 2.76). As further hypothesized (H2), results indicate that
articulating a counterattitudinal referral has a significant positive reliability and validity of our measures led to reasonable results for
all instruments (see Table A1).
effect on the recommender's loyalty (F(1,99) ¼12.69, p o.001,
partial η2 ¼.12, MCG ¼2.25, MTG ¼ 3.06).
3.3.3. Manipulation and realism checks
Again, we ensured that the attitudinal scenario produced a
Table 1 negative attitude toward “Health-Plus” (M(120)¼ 2.05). All groups
MANCOVA results for attitude and loyalty (Study 1).
had similar levels of attitude (F(2,120) ¼.32, p ¼ .73, MCG ¼ 2.02,
Mean value Mean value F-value p-value Partial η2 M10€ ¼1.99, M100€ ¼2.13). Also, we carried out manipulation checks
control group treatment for reward size and the magnitude of dissonance in both referral
(n¼ 50) group (n¼ 49) groups. As intended, participants perceived the value of the large
reward to be significantly greater than the value of the small
Attitude 1.95 2.76 21.82 .000 .19
Loyalty 2.25 3.06 12.69 .001 .12
reward (t(81) ¼  8.94, po .001, M10€ ¼2.61, M100€ ¼5.01). Corre-
sponding with our theoretical framework, the students in the
M. Kuester, M. Benkenstein / Journal of Retailing and Consumer Services 21 (2014) 897–904 901

small reward group felt more uncomfortable giving a counter- study complements previous work on sender-related impacts of
attitudinal referral than participants in the large reward group natural and stimulated WOM (Eggert et al., 2007; Garnefeld et al.,
(t(81) ¼  5.04, p o.001, M10€ ¼2.73, M100€ ¼4.58). We further 2011, 2013).
made sure that students articulated positive WOM in the speech Our results support the idea of Garnefeld et al. (2013) that
bubble. Similar to Study 1, participants in all three experimental participating in a referral program constitutes a public commit-
conditions indicated that they could picture themselves in the ment that evokes consistency effects. By drawing on Festinger's
described situation (M(120) ¼ 3.76). No significant differences (1957) dissonance theory, we confirmed the applicability of the
were found between the realism perceptions of all three groups commitment-consistency framework for counterattitudinal refer-
(F(2,120) ¼.04, p ¼.96, MCG ¼ 3.71, M10€ ¼3.73, M100€ ¼3.83). rals. To retrieve inner consistency, recommenders align their
attitude and subsequent behaviour with their attitude-discrepant
3.3.4. Findings referral. However, we found that the positive impact of articulating
Firstly, we tested whether the replication of the positive effect a counterattitudinal referral depends on the size of the reward.
of giving a counterattitudinal recommendation on the sender's Our findings indicated an inverse incentive effect; in other words,
attitude and loyalty was successful. For this purpose, we merged participation in a referral program has a positive effect on the
the two treatment groups into one group and used the pretest communicator's attitude and loyalty when the rewards are small,
score of participants' attitude toward the health insurance provi- but has no impact when the service provider offers large rewards.
der as a covariate. Again, we found a significant effect of The results thus suggest that larger incentives are ineffective to
giving counterattitudinal referrals on the recommenders' attitude improve unsatisfied customers' attitude and loyalty toward the
(F(1,120) ¼4.95, p o.05, partial η2 ¼ .04, MCG ¼2.17, MTG ¼2.48) and recommended service provider. This notable finding is in line with
loyalty (F(1,120)¼ 8.79, p o.01, partial η2 ¼.07, MCG ¼2.23, Garnefeld et al.'s (2013) speculation that large rewards do not
MTG ¼ 2.77). The results thus provide further support for H1 and enhance attitudes in attitude-discrepant referral behaviour. Also,
H2 (see Table 2). Ryu and Feick (2007) have generally asserted that a recommenda-
The moderating effect of reward size (H3) was tested conduct- tion induced by a large extrinsic incentive may undermine the
ing a separate MANCOVA with all three groups. The analysis positive impact of unrewarded WOM. From a dissonance theory
revealed a major effect both on attitude (F(2,120) ¼5.46, po .01, perspective, a large reward offers an external cause that explains
partial η2 ¼.09, MCG ¼2.17, M10€ ¼2.59, M100€ ¼2.37) and loyalty the counterattitudinal recommendation. Hence, larger induce-
(F(2,120) ¼8.65, p o.001, partial η2 ¼.13, MCG ¼ 2.23, M10€ ¼3.01, ments thoroughly justify participating in a referral campaign
M100€ ¼ 2.51). As postulated in H3, Bonferroni tests indicated a despite being dissatisfied. Recommenders perceive the incentive
significant difference between the control group and the small- as a sufficient motivation that has evoked their opportunistic
rewarded treatment group in terms of their attitude (po .01) and behaviour. As customers know that their referral was merely
loyalty (po .001), whereas the attitude (p¼ 1.0) and loyalty driven by the large reward, they do not experience a discomfort
(p ¼.67) between the control group and the large-rewarded group and do not have to align their attitude and their loyalty with their
did not significantly differ. Compared with the large-rewarded attitude-discrepant recommendation.
condition, articulating a counterattitudinal referral for a small In addition, making an incentivized recommendation may result
reward had a positive effect on the communicator's attitude in a trade-off between the motives of self-interest and maintenance of
(p o.05) and loyalty (po.05). In line with H3, we found articulat- the relationship with the recipient of the referral (Kornish and Li,
ing a counterattitudinal referral had a positive effect on the 2010; Wirtz et al., 2013). As the recommender has a personal
recommender's attitude and loyalty for referral programs with relationship with the receiver, rewarded referrals are not just a purely
small rewards, but not for campaigns with large incentives. Table 3 commercial communication (Wirtz et al., 2013). They always entail a
summarizes the results. social risk (Tuk et al., 2009), which refers to the possible negative
consequences arising from the recipient's reaction (Wangenheim and
Bayón, 2004). Customers therefore balance the value of the reward to
4. Discussion and implications themselves with the potential social risk of negatively affecting the
relationship (Ryu and Feick, 2007). With opportunistic referrals, we
Past research has found that participating in a customer referral expect recommenders to perceive high social costs. The relationship
program positively affects the behavioural loyalty of the recom- would probably be jeopardized if the receiver found out the ulterior
mender, and that campaigns with larger rewards also strengthen motives of the referral and the recommender's real attitude toward
the referring customer's attitudinal loyalty (Garnefeld et al., 2013). the provider. However, a large incentive should be perceived as an
However, previous studies have overlooked the possible impact on appropriate compensation for the potential social risk. In the case of a
unsatisfied customers who recommend their service providers in large reward, the personal financial benefits thus should outbalance
order to obtain the reward. The findings of this research suggest the social costs. The referring customer therefore does not feel a need
that articulating attitude-discrepant referrals within a referral to make a positive impression on the receiver by being consistent
reward campaign has a positive effect on the recommender's with the articulated referral.
attitude and loyalty toward the recommended service provider. By showing that small-incentivized referral campaigns improve
By providing the first empirical evidence regarding the effects of opportunistic customers' attitude and loyalty toward the service
giving a counterattitudinal referral on the referring customer, this provider, our study recognizes the meaning of referral programs as
means for enhancing and retaining relationships with existing
customers. Our findings thereby reinforce the recent idea that
Table 2 customer referral programs are not just a tool to acquire new
MANCOVA results for attitude and loyalty (Study 2).
customers (e.g., Kornish and Li, 2010; Kumar et al., 2010; Schmitt
Mean value Mean value F-value p-value Partial η2 et al., 2011; Xiao et al., 2011), but also act as an instrument for
control group treatment customer relationship management (Garnefeld et al., 2011, 2013;
(n¼39) group (n¼ 81) Reinartz et al., 2004; Ryu and Feick, 2007). Since dissatisfied
customers evaluate the service provider in a more favourable
Attitude 2.17 2.48 4.95 .028 .04
Loyalty 2.23 2.77 8.79 .004 .07
manner after participating in small-rewarded campaigns, these
programs may particularly help to increase consumers' satisfaction
902 M. Kuester, M. Benkenstein / Journal of Retailing and Consumer Services 21 (2014) 897–904

Table 3
MANCOVA results for the moderating effect of reward size.

Mean value control Mean value small Mean value large F-value p-value Partial η2 Bonferroni test (p-value)
group (n ¼39) reward group reward group
(n ¼41) (n ¼40) Control vs. Small Control vs. Large Small vs. Large

Attitude 2.17 2.59 2.37 5.46 .005 .09 .006 1.000 .054
Loyalty 2.23 3.01 2.51 8.65 .000 .13 .000 .673 .017

with the provider. Marketing managers are therefore encouraged 5. Limitations and further research
to use member-get-member-campaigns as a supplementary
instrument to deal with unsatisfied customers. However, providers As with any other study that focuses exclusively on a singly
should be aware that targeting member-get-member-programs to service context and uses student samples, limitations persist
unsatisfied customers might undermine the additional goal to regarding the generalizability of the results. A homogeneous
attract new customers. First, receivers' responses to the referral student population makes the situational variance controllable,
might be adversely affected. If the recipient found out that the and minimizes the likelihood of making a Type II error (Calder
referral was merely motivated by the reward, the referred custo- et al., 1981); however, it may also result in a drift toward stronger
mer would question the sincerity of the recommender and effects than would be found in the general population (Brown and
possibly not subscribe to the provider. According to the findings Stayman, 1992). In order to confirm the robustness of the findings,
of Verlegh et al. (2013), rewarding both the recommender and the further research should replicate this study in different service
receiver can eliminate this negative effect. Second, the relationship settings with non-student subjects.
between the newly acquired customer and the provider would be wIn our experiments, participants were asked to project
harmed if the receiver detected the ulterior motives after signing themselves into a fictitious situation. A few authors assume low
up for a contract. Further studies should address this issue in order external validity for scenario-based experiments (e.g., Aronson
to clarify the possible trade-off between the goals of enhancing et al., 1990). However, the realism check showed that subjects
customers' satisfaction and acquiring new customers. Besides could imagine the scenarios. But for higher external validity, the
actively using referral campaigns to increase customers' satisfac- results should be validated in field experiments with real referrals.
tion, firms can take advantage of opportunistic referral behaviour Based on existent referral campaigns in the health insurance
in general. As opportunism is a prevalent accompaniment of industry, we operationalized the size for small and large rewards
referral reward campaigns, providers can also indirectly profit with €10 and €100 respectively. The sizes of referral rewards may
from the favourable side effects of self-interested behaviour on the vary in different service industries. Additional research is therefore
recommender's attitude and loyalty. necessary to provide the adequate sizes of smaller rewards for
The findings of this study may further help managers design different types of services. We further argue that referral reward
referral reward programs. As the positive effects on dissatisfied programs can be a tool to enhance unsatisfied customers' attitude
customers only arise in a customer referral program with smaller and loyalty. However, we are not familiar with what reward sizes
rewards, service firms would be well advised to launch member-get- actually evoke the opportunistic referral behaviour of customers.
member-campaigns with small rewards. This implication corresponds In order to complete the findings, future research should investi-
to the findings of Ryu and Feick (2007) concerning new customer gate if smaller rewards are sufficient to provoke opportunism. We
acquisition. Their study showed that small rewards are sufficient to presume that smaller rewards will contribute to reduce opportu-
increase referral likelihood, and that an increase in reward size does nistic WOM behaviour. Regarding the impact of incentives on
not lead to a rise in likelihood. According to the findings of Garnefeld unsatisfied customers, we expect an opposing effect: while custo-
et al. (2013), only larger rewards strengthen both the communicators' mers' willingness to recommend an unsatisfactory service provi-
attitude and behavioural loyalty when customers recommend satis- der may decline along with smaller referral rewards, the positive
factory service providers. However, their study revealed that small effects of articulating a counterattitudinal referral only arise in
rewards are adequate to increase senders' loyalty. By offering small referral reward programs with smaller rewards. Thus, further
rewards, service providers cannot do anything wrong: they may gain analyses should focus on this optimizing issue in order to find
new customers, enhance satisfied clients' loyalty, and strengthen the optimal reward size for unsatisfied customers. In this context,
unsatisfied customers' attitude and loyalty. research could pay specific attention to the tie strength between
Finally, our research contributes to debilitate prior scepticism the communicator and the recipient. Opportunistic WOM beha-
about referral campaigns. Several authors doubt the benefits of viour was found to occur more frequently in communications to
stimulated referrals (e.g., Trusov et al., 2009; Van den Bulte, 2010). weak ties (Wangenheim, 2003). According to Ryu and Feick
Referral reward programs are deemed to be very cost effective (2007), rewards increase referral likelihood more for weak than
(Schmitt et al., 2011; Xiao et al., 2011) and prone to the opportu- for strong ties. In the same vein, we assume that opportunism in
nistic behaviour of customers (Schmitt et al., 2011; Wirtz and referral reward programs is more likely to become an issue in
Chew, 2002). By demonstrating that small-rewarded counteratti- weak tie relationships because the recommender may be less
tudinal referrals strengthen referring customers' attitude and concerned about the negative impact on the personal relationship
loyalty, we show that referral campaigns can pay off; not only in with the recipient.
terms of newly acquired customers, but also in terms of the In practice, many firms use non-financial incentives, such as credit
referring customers, even when those programs are abused by and gifts, to encourage their customers to provide referrals (Kornish
unsatisfied customers. Our results thus support the reasoning of and Li, 2010; Wirtz and Chew, 2002). In some of these programs, the
Schmitt et al. (2011) that the benefits of referral reward programs recommender even has to make an extra payment to receive the non-
outbalance the harmful side effects such as the abuse by oppor- financial reward. Service providers might provide large non-monetary
tunistic customers. Since firms can even profit from self-interested rewards to enhance unsatisfied customers' willingness to participate
customers with regards to the bonding effect, a stronger focus on in the referral campaign. We surmise that the referrers' additional
referral reward programs is worth considering. payment may induce cognitive dissonance irrespective of the large
M. Kuester, M. Benkenstein / Journal of Retailing and Consumer Services 21 (2014) 897–904 903

reward. Hence, articulating a counterattitudinal referral might have a courses, etc.). Also, you had to pay an additional fee for your
positive effect on the recommender's attitude and loyalty even within insurance coverage.
referral reward programs with larger non-financial rewards. For a 2. Manipulation of a referral
better understanding about the impact of member-get-member- a. Small reward
campaigns, follow-up studies using non-monetary incentives with Imagine meeting a friend at a coffee bar. The friend tells you
extra payment would be useful. that she or he is looking for a new health insurance
provider. While you talk about health insurance, you
remember that you recently received a flyer from your
Acknowledgements insurance provider, “Health-Plus”, with a request to recom-
mend it to friends. If participating in the member-get-
member-campaign, you would receive a reward of €10 for
The authors would like to thank Yvonne Strelow for her
every successfully acquired customer as a thank you. There-
research assistance.
upon, you recommend your health insurance provider
“Health-Plus” to the friend.
Please articulate your referral in the speech bubble!
Appendix b. Large reward
Imagine meeting a friend at a coffee bar. The friend tells you
Study 1 that she or he is looking for a new health insurance
provider. While you talk about health insurance, you
1. Manipulation of a negative attitude remember that you recently received a flyer from your
Please imagine that you switched to the statutory health insurance provider, “Health-Plus”, with a request to recom-
insurance provider “Health-Plus” three years ago. Since your mend it to friends. If participating in the member-get-
signing with “Health-Plus”, most contact has been rather member-campaign, you would receive a reward of €100
displeasing (signing up without personal consulting service, for every successfully acquired customer as a thank you.
no information about additional services, no subsidized health Thereupon, you recommend your health insurance provider
courses, etc.). Also, you had to pay an additional fee for your “Health-Plus” to the friend.
insurance coverage.. Please articulate your referral in the speech bubble!
2. Manipulation of a referral
Imagine meeting a friend at a coffee bar. The friend tells you that
she or he is looking for a new health insurance provider. While you
talk about health insurance, you remember that you recently
received a flyer from your insurance provider, “Health-Plus”, with
a request to recommend it to friends. If participating in the I am looking for a
member-get-member-campaign, you would receive a monetary new health insurance
provider !
reward for every successfully acquired customer as a thank you.
Thereupon, you recommend your health insurance provider
“Health-Plus” to the friend.
Please articulate your referral in the speech bubble!

Study 2

1. Manipulation of a negative attitude

Please imagine that you switched to the statutory health
insurance provider “Health-Plus” three years ago. Since your
signing with “Health-Plus”, most contact has been rather
displeasing (signing up without personal consulting service, Fig. A1. Manipulations for Studies 1 and 2. (The image has been downloaded from
no information about additional services, no subsidized health ¼ 6&ver¼ 14&app ¼powerpnt.exe.)

Table A1
Measures and validation.

Construct and items Factor loading Explained Item-to-total Cronbach's α

variance (%) correlation

Study Study Study Study Study Study Study Study

1 2 1 2 1 2 1 2

Attitude 84.6 81.3 .91 .89

“Health-Plus” is very negative/very positive. .94 .93 .86 .82
“Health-Plus” is very unfavourable/very favourable. .93 .89 .83 .75
“Health-Plus” is very bad/very good. .89 .89 .76 .75

Loyalty 55.5 57.1 .71 .74

I am likely to be insured at “Health-Plus” in a year. .73 .72 .46 .49
I am likely to make negative comments about “Health-Plus” to my friends and family.a .62 .77 .40 .55
If my health insurance provider “Health-Plus” were to raise my premium, I would still continue to be .83 .83 .62 .65
a customer.
If a competing health insurance provider were to offer a lower premium, I would switch.a .78 .70 .57 .48

Items were reverse scaled.
904 M. Kuester, M. Benkenstein / Journal of Retailing and Consumer Services 21 (2014) 897–904

References Kiesler, C.A., 1971. The Psychology of Commitment: Experiments Linking Behavior
to Belief. Academic Press, New York.
Kiesler, C.A., Pallak, M.S., 1976. Arousal properties of dissonance manipulations.
Aronson, E., Ellsworth, P.C., Carlsmith, J.M., Gonzales, M.H., 1990. Methods of Psychol. Bull. 83 (6), 1014–1025.
Research in Social Psychology. Mc Graw-Hill Publishing, New York. Kornish, L.J., Li, Q., 2010. Optimal referral bonuses with asymmetric information:
Biyalogorsky, E., Gerstner, E., Libai, B., 2001. Customer referral management: firm-offered and interpersonal incentives. Mark. Sci. 29 (1), 108–121.
optimal reward programs. Mark. Sci. 20 (1), 82–95. Kumar, V., Petersen, J.A., Leone, R.P., 2010. Driving profitability by encouraging
Blazevic, V., Hammedi, W., Garnefeld, I., Rust, R.T., Keiningham, T., Andreassen, T.W., customer referrals: who, when, and how. J. Mark. 74 (5), 1–17.
Donthu, N., Carl, W., 2013. Beyond traditional word-of-mouth. An expanded Laczniak, R.N., DeCarlo, T.E., Ramaswami, S.N., 2001. Consumers' responses to
model of customer-driven influence. J. Serv. Manag. 24 (3), 294–313. negative word-of-mouth communication: an attribution theory perspective.
Brehm, J.W., Cohen, A.R., 1962. Explorations in Cognitive Dissonance. Wiley, New
J. Consum. Psychol. 11 (1), 57–73.
York. Losch, M.E., Cacioppo, J.T., 1990. Cognitive dissonance may enhance sympathetic
Brown, S.P., Stayman, D.M., 1992. Antecedents and consequences of attitude toward
tonus, but attitudes are changed to reduce negative affect rather than arousal.
the ad: a meta-analysis. J. Consum. Res. 19 (1), 34–51.
J. Exp. Soc. Psychol. 26 (4), 289–304.
Calder, B.J., Phillips, L.W., Tybout, A.M., 1981. Designing research for application.
Nyer, P.U., Dellande, S., 2010. Public commitment as a motivator for weight loss.
J. Consum. Res. 8 (2), 197–207.
Psychol. Mark. 27 (1), 1–12.
Carlsmith, J.M., Collins, B.E., Helmreich, R.L., 1966. Studies in forced compliance: the
Nyer, P.U., Gopinath, M., 2005. Effects of complaining versus negative word of
effect of pressure for compliance on attitude change produced by face-to-face
mouth on subsequent changes in satisfaction: the role of public commitment.
role playing and anonymous essay writing. J. Pers. Soc. Psychol. 4 (1), 1–13.
Psychol. Mark. 22 (12), 937–953.
Cialdini, R.B., 1971. Attitudinal advocacy in the verbal conditioner. J. Pers. Soc.
Olson, J.M., Stone, J., 2005. The influence of behavior on attitudes. In: Albarracín, D.,
Psychol. 17 (3), 350–358.
Johnson, B.T., Zanna, M.P. (Eds.), The Handbook of Attitudes. Lawrence Erlbaum
Cialdini, R.B., 2009. Influence: Science and Practice. Pearson Education, Boston.
Associates, Mahwah, pp. 223–271.
Cialdini, R.B., Goldstein, N.J., 2004. Social influence: compliance and conformity.
Pallak, M.S., Cummings, W., 1976. Commitment and voluntary energy conservation.
Annu. Rev. Psychol. 55, 591–621.
Pers. Soc. Psychol. Bull. 2 (1), 27–31.
Cioffi, D., Garner, R., 1996. On doing the decision: effects of active versus passive
Pomerantz, E.M., Chaiken, S., Tordesillas, R.S., 1995. Attitude strength and resistance
choice on commitment and self-perception. Pers. Soc. Psychol. Bull. 22 (2),
process. J. Pers. Soc. Psychol. 69 (3), 408–419.
Reinartz, W., Krafft, M., Hoyer, W.D., 2004. The customer relationship management
Cooper, J., Carlsmith, K.M., 2001. Cognitive dissonance. In: Smelser, N.J., Baltes, P.B.
process: its measurement and impact on performance. J. Mark. Res. 41 (3),
(Eds.), International Encyclopedia of the Social & Behavioral Sciences, vol. 3.
Elsevier, Amsterdam, pp. 2112–2114.
Rosen, E., 2001. The Anatomy of Buzz: Creating Word-of-Mouth-Marketing.
Davidow, M., 2003. Have you heard the word? The effect of word of mouth on
HarperCollins, London.
perceived justice, satisfaction and repurchase intentions following complaint
Ryu, G., Feick, L., 2007. A penny for your thoughts: referral reward programs and
handling. J. Consum. Satisf. Dissatisf. Complain. Behav. 16, 67–80.
referral likelihood. J. Mark. 71 (1), 84–94.
Eagly, A.H., Chaiken, S., 1993. The Psychology of Attitudes. Harcourt Brace
Schmitt, P., Skiera, B., Van den Bulte, C., 2011. Referral programs and customer
Jovanovich, Fort Worth.
Eagly, A.H., Chaiken, S., 1995. Attitude strength, attitude structure, and resistance to value. J. Mark. 75 (1), 46–59.
change. In: Petty, R.E., Krosnick, J.A. (Eds.), Attitude Strength: Antecedents and Simmons, C.J., Becker-Olsen, K.L., 2006. Achieving marketing objectives through
Consequences. Erlbaum, Mahwah, pp. 413–432. social sponsorships. J. Mark. 70 (4), 154–169.
Eggert, A., Helm, S., Garnefeld, I., 2007. Kundenbindung durch Weiterempfehlung? Söderlund, M., Rosengren, S., 2007. Receiving word-of-mouth from the service
Eine experimentelle Untersuchung der Wirkung positiver Kundenempfehlun- customer: an emotion-based effectiveness assessment. J. Retail. Consum. Serv.
gen auf die Bindung des Empfehlenden [Loyalty through referrals? An experi- 14 (2), 123–136.
mental study about the impact of positive customer referrals on the referrer's Stone, J., Fernandez, N.C., 2008. How behavior shapes attitudes. In: Crano, W.D.,
loyalty]. Mark. ZFP 29 (4), 233–245. Prislin, R. (Eds.), Attitudes and Attitude Change. Psychology Press, New York,
Elkin, R.A., Leippe, M.R., 1986. Physiological arousal, dissonance, and attitude pp. 313–334.
change: evidence for a dissonance-arousal link and a “Don't Remind Me” effect. Tax, S.S., Chandrashekaran, M., Christiansen, T., 1993. Word-of-mouth in consumer
J. Pers. Soc. Psychol. 51 (1), 55–65. decision-making: an agenda of research. J. Consum. Satisf. Dissatisf. Complain.
Fazio, R.H., Zanna, M.P., Cooper, J., 1977. Dissonance and self-perception: an Behav. 6, 74–80.
integrative view of each theory's proper domain of application. J. Exp. Soc. Trusov, M., Bucklin, R.E., Pauwels, K., 2009. Effects of word-of-mouth versus
Psychol. 13 (5), 464–479. traditional marketing: findings from an internet social networking site. J. Mark.
Festinger, L., 1957. A Theory of Cognitive Dissonance. Stanford University Press, 73 (5), 90–102.
Stanford. Tuk, M.A., Verlegh, P.W.J., Smidts, A., Wigboldus, D.H.J., 2009. Sales and sincerity:
Frey, D., Gaska, A., 2001. Die Theorie der kognitiven Dissonanz [The cognitive the role of relational framing in word-of-mouth marketing. J. Consum. Psychol.
dissonance theory]. 2nd ed. In: Frey, D., Irle, M. (Eds.), Theorien der Sozialp- 19 (1), 38–47.
sychologie: Kognitive Theorien [Theories of Social Psychology: Cognitive Van den Bulte, C., 2010. Opportunities and challenges in studying customer
Theories], vol. 2. Huber, Bern, pp. 275–326. networks. In: Wuyts, S., Dekimpe, M.G., Gijsbrechts, E., Pieters, R. (Eds.), The
Ganesh, J., Arnold, M.J., Reynolds, K.E., 2000. Understanding the customer base of Connected Customer: The Changing Nature of Consumer and Business Markets.
service providers: an examination of the differences between switchers and Routledge, New York, pp. 7–35.
stayers. J. Mark. 64 (3), 65–87. Verlegh, P.W.J., Ryu, G., Tuk, M.A., Feick, L., 2013. Receiver responses to rewarded
Garnefeld, I., Muenkhoff, E., Hogreve, J., Eggert, A., 2009. Unintended effects of referrals: the motive inferences framework. J. Acad. Mark. Sci. 41 (6), 669–682.
referral reward programs: exploring the impact of rewards on opportunistic Wang, T.H., Katzev, R.D., 1990. Group commitment and resource conservation: two
behavior. In: Proceedings of the 38th EMAC Conference, Nantes, p. 210. field experiments on promoting recycling. J. Appl. Soc. Psychol. 20 (4), 265–275.
Garnefeld, I., Helm, S., Eggert, A., 2011. Walk your talk: an experimental investiga- Wang, X., 2011. The effect of inconsistent word-of-mouth during the service
tion of the relationship between word of mouth and communicators' loyalty. encounter. J. Serv. Mark. 25 (4), 252–259.
J. Serv. Res. 14 (1), 93–107. Wangenheim, F.v., 2003. Opportunism in Interpersonal Exchange: When Dissatis-
Garnefeld, I., Eggert, A., Helm, S.V., Tax, S.S., 2013. Growing existing customers' faction is Followed by Positive Word-of-Mouth (Working Paper). University of
revenue streams through customer referral programs. J. Mark. 77 (4), 17–32. Dortmund, Dortmund.
Gopinath, M., Nyer, P.U., 2009. The effect of public commitment on resistance to Wangenheim, F.v., Bayón, T., 2004. The effect of word of mouth on services
persuasion: the influence of attitude certainty, issue importance, susceptibility switching. Eur. J. Mark. 38 (9/10), 1173–1185.
to normative influence, preference for consistency and source proximity. Int. J. Wirtz, J., Chew, P., 2002. The effects of incentives, deal proneness, satisfaction and
Res. Mark. 26 (1), 60–68. tie strength on word-of-mouth behaviour. Int. J. Serv. Ind. Manag. 13 (2),
Greenwald, A.G., Carnot, C.G., Beach, R., Young, B., 1987. Increasing voting behavior 141–162.
by asking people if they expect to vote. J. Appl. Psychol. 72 (2), 315–318. Wirtz, J., Orsingher, C., Chew, P., Tambyah, S.K., 2013. The role of metaperception on
Helm, S., 2000. Kundenempfehlungen als Marketinginstrument [Customer referrals the effectiveness of referral reward programs. J. Serv. Res. 16 (1), 82–98.
as marketing tool]. Deutscher Universitaets-Verlag, Wiesbaden. Xiao, P., Tang, C.S., Wirtz, J., 2011. Optimizing referral reward programs under
Jin, L., Huang, Y., 2013. When giving money does not work: the differential effects of impression management considerations. Eur. J. Oper. Res. 215 (3), 730–739.
monetary versus in-kind rewards in referral reward programs. Int. J. Res. Mark. Zanna, M.P., Cooper, J., 1974. Dissonance and the pill: an attribution approach to
31 (1), 107–116. studying the arousal properties of dissonance. J. Pers. Soc. Psychol. 29 (5),
Katzev, R.D., Pardini, A.U., 1987. The comparative effectiveness of reward and 703–709.
commitment approaches in motivating community recycling. J. Environ. Syst. Zanna, M.P., Higgins, E.T., Taves, P.A., 1976. Is dissonance phenomenologically
17 (2), 93–113. aversive? J. Exp. Soc. Psychol. 12 (6), 530–538.