1 Introduction

Although dissatisfied customers are expected to voice their problems to the firms, the majority of customers stay silent about service failures (Colgate and Hedge 2001; Stephens and Gwinner 1998). An unvoiced complaint is considered an “opportunity cost” because the firm misses an opportunity that allows them to identify the source of the service problem and provide a service recovery for customer retention (Bell et al. 2004; Fornell and Wernerfelt 1987). Non-complainers may become detrimental to the companies because they often switch to another company and advise others to do the same (Tax et al. 1998; Voorhees et al. 2006). However, very little research exists that has investigated non-complainers’ reactions to the service failures.

Complaint management literature consistently suggests that customers who have had a satisfying complaint handling are more likely to return to the service provider than customers who have had a dissatisfying one (Maxham and Netemeyer 2002). However, return intentions of non-complainers are not clear. Do they forget about the dissatisfying incident and return? Or do they remain dissatisfied, and thus, do not return? This either optimistic or pessimistic bet on non-complainers’ return intentions does not provide realistic answers for the service firms. In fact, previous studies have shown mixed results for the non-complainers’ return intentions. Some studies suggest that non-complainers are more likely to return than complainers (e.g., Voorhees et al. 2006) while others report the opposite (e.g., TARP 1999). These mixed results may come from a lack of understanding in regard to the nature of dissatisfaction responses from non-complainers.

Some people forgo complaining because they want to give the service provider another chance. Others disregard complaining because they do not care about the company and for them it is not worth taking any action. In the organizational behavior literature, these two types of passive dissatisfaction responses are termed patience and neglect, respectively (Hagedoorn et al. 1999). Investigating these two dissatisfaction responses in the consumer behavior context can provide useful insights for understanding non-complainers reactions to service failures.

Although customers are not restricted to a single complaining behavior but may engage in multiple complaining behaviors (e.g., complaining to the service provider, spreading negative word-of-mouth, and engaging in third-party actions) (Day et al. 1981; Singh 1988), this study focuses on the direct complaint to the service provider or the absence of it at the time of a service failure. When a customer encounters a service failure, there will either be a complaint or no complaint. If the customer complains, the outcome of the service recovery will result in either a satisfied or a dissatisfied customer. If, however, the customer does not complain, his or her no-complain decision can be further described as patience or neglect.

This study investigates customers’ dissatisfaction responses to a service failure. Three dissatisfaction responses (complaint, patience, and neglect) are distinguished by related variables: attitude toward complaining, emotional bonding, and criticality of the service failure. In addition, return intentions are examined using two groups of complainers, recovery satisfied and recovery dissatisfied, and two groups of non-complainers, patience and neglect.

From a theoretical standpoint, this study identifies two different non-complainers’ dissatisfaction responses, therefore contributing to the consumer complaint behavior literature and expanding its taxonomy. In addition, findings of this study provide clarification for the mixed findings in past research in regard to non-complainers’ return intentions. From a managerial standpoint, this study provides service organizations with a better understanding of the silent dissatisfied customers and emphasizes relationship building with customers.

The article is organized as follows: first, previous research on consumer complaint behavior and dissatisfaction responses is reviewed. Then, the hypotheses that distinguish dissatisfaction responses by antecedents and return intentions are developed. Next, the methods and results from the auto repair and medical service contexts are discussed. Finally, the implications for the researchers and practitioners are also discussed.

2 Literature review

2.1 Consumer complaining behavior

Hirschman (1970) suggests that people in a deteriorating relationship have three basic response options: leave the relationship (exit), talk about the problem (voice), or quietly remain in the relationship (loyalty). Day and Landon (1977) introduce a hierarchical model of consumer complaint behavior (CCB) taxonomy, which suggests that dissatisfied consumers would either take “action” or “no-action”. If action is taken, it is labeled as either public (e.g., redress seeking complaint, third-party actions) or private action (e.g., negative word-of-mouth, personal boycott of the seller). Conversely, if no action is taken, it is described as “forget about the incident and do nothing at all.” Based on these two frameworks, Singh (1988) empirically validates three categories of CCB: (1) voice, reflecting actions directed toward the seller; (2) private, involving negative word-of-mouth (WOM) and exit; (3) third party, relating to actions involving external agencies such as better business bureau and legal actions.

Although the complainers’ responses have been extensively investigated in the CCB literature, the non-complainers’ responses are not clearly theorized. For example, Hirschman’s “loyalty” implies not complaining but being patient with the negative experience. Similarly, Day and Landon’s “no action” response is understood as a complete void of all behavioral actions including switching behavior, thus assuming that people who take no-action will return. Other researchers suggest that “no action” may not necessarily mean complete inaction from the customers’ point of view. For example, Zeelenberg and Pieters (2004) identified “inertia” where some customers remain passive and thus fail to take action in response to a service failure. Although “inertia” customers describe their responses as doing nothing, similar to no action, they tend to engage in other private actions such as negative WOM and switch (Zeelenberg and Pieters 2004). Due to its unclear conceptualization, the “no action” response sometimes disappears or is mixed in with other CCB categories (e.g., Hansen et al. 1997; Singh 1988).

2.2 Neglect and patience

Extending Hirschman’s exit, voice, and loyalty framework, Rusbult et al. (1982) suggest additional passive dissatisfaction response category—neglect. According to Rusbult et al. (1982), people may actively deal with a problem by expressing concerns in order to revive the relationship, or making a decision to remove themselves from the relationship (via voice or exit, respectively). Alternatively, people may passively deal with a problem by being patient with the situation or by letting the relationship to atrophy (via loyalty or neglect, respectively). Therefore, they suggest two types of passive dissatisfaction responses that do not involve immediate actions.

Loyalty (or patience) is described as taking no action and remaining with the firm by believing that the situation will improve (Geyskens and Steenkamp 2000; Hagedoorn et al. 1999; Hirschman 1970). Rusbult et al. (1982, 1988) describe loyalty as a passive but constructive behavior because it is directed at improving the relationship by being quietly supportive. This concept has been given different labels, such as “stay-silent” (Kolarska and Aldrich 1980) and “patience” (Hagedoorn et al. 1999). On the other hand, neglect is described as a dispassionate and psychological withdrawal whereby the individual becomes apathetic toward the relationship and is not willing to communicate the dissatisfaction (Oliver 1997; Rusbult et al. 1988). In a channel relationship study, Ping (1993) describes neglect as a type of “emotional exiting” where people do not care about the partner firm and let the relationship deteriorate. Kolarska and Aldrich (1980) also note that customer dissatisfaction may lead to apathy when exit is impeded and voice is ineffective.

Based on the review of past research, this study posits that the non-complainers can be understood via two distinct responses: patience and neglect. Patience is defined as not complaining because of willingness to give the service provider another chance by hoping that the negative situation will improve in the future. Neglect, on the other hand, is defined as not complaining due to being indifferent about the company and believing that taking any action is not worthwhile.

2.3 Customer dissatisfaction responses

As shown in Fig. 1, dissatisfied customers can either complain or not complain to the service provider. If customers decide not to complain, they are labeled either patience or neglect reflecting the different nature of each response. Once customers complain, they are further categorized into recovery satisfied and recovery dissatisfied based on the service recovery outcome. In the following section, the selected antecedents that distinguish patience and neglect from complaint are discussed.

Fig. 1
figure 1

Customer dissatisfaction responses after a dissatisfying service experience

Full size image

2.4 Antecedents of customer dissatisfaction responses

CCB literature suggests that customers’ decision to complain or not is a complicated process triggered by dissatisfaction and influenced by individual and situational factors (Stephens and Gwinner 1998). Early CCB researchers tried to link complaining with demographic variables, such as education and income (Grønhaug and Zaltman 1981; Morganosky and Buckley 1987; Warland et al. 1975). Others suggest that complaining behavior is better predicted by attitudinal or personality variables such as attitude toward complaining, assertiveness, self-confidence, and Machiavellianism (Bodey and Grace 2007; Singh and Wilkes 1996; Swanson 2007). Although individuals have a varying degree of general tendency to complain, researchers suggest that the complaint decision is influenced by situational factors, such as, attribution (Singh and Wilkes 1996), switching convenience (Keaveney 1995), and perceived costs and benefits of engaging in complaining behaviors (Singh and Wilkes 1996). Thus, research findings on what influences complaining behaviors are complicated by varied results.

While past research often makes a binary distinction between complaint and no-complaint (e.g., Bodey and Grace 2007; Day and Landon 1977; Grønhaug and Zaltman 1981), this study examines three responses after a dissatisfying experience: patience, neglect and complaint. In order to distinguish the three responses, this research focuses on the attitude toward complaining, criticality of the service failure, and emotional bonding.

Attitude toward complaining refers to an individual’s predisposition toward complaining to service providers (Singh and Wilkes 1996) and it is found to be an important predictor in CCB (Blodgett and Anderson 2000; Chebat et al. 2005). Some people are assertive and will complain whenever they are dissatisfied with a product or service, while others are reluctant to complain even when highly dissatisfied (Swanson 2007). Since taking no action after a service failure may well reflect an individual’s general disinclination when it comes to complaining, it is expected that both patience and neglect responses are driven by the low attitude toward complaining while the complaint response is more likely motivated by high attitude toward complaining.

Hypothesis 1

Attitude toward complaining leads customers to choose complaint over patience or neglect.

Criticality of the service failure can be described as the intensity of the service failure that customer experiences. It serves an important role in determining customers’ response to a service failure and recovery efforts (Hess et al. 2003; Webster and Sundaram 1998). Folkes et al. (1987) suggest that importance of the service is positively related to customers’ anger following failures. As service importance increases, so does the magnitude of the gains and losses experienced by customers. Likewise, the severity of the failure affects the customer service exchange evaluation. The more serious the service failure is, the more likely the customers are to voice their dissatisfaction directly to the service providers (Oliver and Swan 1989). Thus, the following hypothesis is proposed:

Hypothesis 2

Criticality of the service failure will lead customers to choose complaint over patience or neglect.

Emotional bonding Previous studies suggest that relationship bonded customers tend to be generous with a service failure and less likely to complain (Tax et al. 1998). For example, Dewitt and Brady (2003) indicate that high rapport customers consider “not complaining” in the service failure situation as a “reward” for the personal bond between the customer and the service provider. Similarly, Mittal et al. (2008) suggest that strongly relationship bonded customers are less likely to complain because they fear negative consequences for the relationship with the service provider. These studies suggest that emotional bonding might reduce customers’ willingness to complain to the service provider. Since patience response reflects emotional attachment while neglect response reflects emotional withdrawal, the following hypotheses are proposed:

Hypothesis 3a

Customers who have a high emotional bond with the service provider are likely to choose patience over complaint.

Hypothesis 3b

Customers who have a low emotional bond with the service provider are likely to choose neglect over complaint.

2.5 Return intentions

For complainers, return intentions are strongly influenced by their perceptions of service recovery (e.g., Blodgett et al. 1997; Maxham and Netemeyer 2002). According to the equity theory, as customers’ inputs-to-outcomes ratio decreases, they become more likely to see social exchanges in a positive light (Adams 1963). Thus, it is rational to predict that the customers who receive a satisfactory service recovery to their complaints are more likely to return than the customers who receive a dissatisfactory service recovery because the latter suffer from a double deviation (Maxham and Netemeyer 2002). However, research findings on non-complainers’ return intentions are uncertain. TARP (1999) study indicated that non-complainers are less likely to repurchase than complainers regardless of the recovery outcome, whether the problem is resolved or not resolved. However, other researchers found the opposite pattern. For example, Voorhees et al. (2006) suggest that non-complainers show higher repurchase intentions than complainers who have not received a service recovery or have received a dissatisfactory recovery. Blodgett and Anderson (2000) also suggest that some non-complainers intend to keep fully patronizing the service provider, despite their dissatisfaction. These contrasting results make it difficult to determine non-complainers’ return intentions as compared with complainers.

These mixed results can be explained better by examining the non-complainers via two response types: patience and neglect. Patience response customers forgive the dissatisfying service experience by hoping the situation improves, while neglect customers are indifferent toward the dissatisfying experience because they do not care about the company. Therefore, these two types of non-complainers may result in a different level of return intentions. More specifically, the patience response customers’ return intentions will be higher than neglect response customers’ because patience implies commitment to the company. In addition, patience response customers will show higher return intentions than complainers with dissatisfactory recoveries because they are more likely to forgive the company than the customers who have experienced a double deviation. However, neglect response customers may not exhibit higher return intentions than complainers with dissatisfactory recoveries because neglect implies a lack of commitment and emotional bonding. Therefore, the following hypotheses are proposed.

Hypothesis 4

Return intentions are higher for patience customers than neglect customers among non-complainers.

Hypothesis 5

Return intentions are higher for patience customers than complainers with dissatisfactory recoveries.

Return intentions may vary by service type due to industry competition or the attraction of competitors (Roos et al. 2004). When consumers have a great deal of choice within the product or service category, switching brands or service providers becomes relatively easier than voicing dissatisfaction (Singh 1990). In contrast, when the customer’s alternatives are limited, dissatisfaction does not easily prompt switching (Singh 1990). For this study, auto-repair and medical services are chosen due to their different characteristics. For example, medical care services often involve high customer contact and customization, while auto-repair services are less personal (Bowen 1990). More importantly, switching is not readily available in medical services but it is easy to switch to another service provider in auto-repair services. Thus, this study hypothesizes the following:

Hypothesis 6

Return intentions are higher for medical services than auto-repair services.

3 Method

3.1 Data collection

The current study used a critical incident survey (e.g., Voorhees et al. 2006). First, respondents were asked to recall one of their own most recent dissatisfying experiences, describe the incident and rate the criticality of the problem. Respondents were then asked to indicate whether they complained to the service provider or not at that time. For those who complained, perceptions of the service recovery were asked. For those who did not complain, patience and neglect descriptions were provided for them to indicate the type of their “no complaint” response. All respondents were asked to rate to a series of scaled items that assess emotional bonding with the company prior to the incident and return intentions after the incident. Finally, demographic data such as gender and age were collected. The questionnaire was pre-tested with a convenient sample of 65 undergraduate students before its administration. Following the pre-test, a few minor changes were made to the layout of the questionnaire to facilitate ease of completion.

The population of interest is consumers who have had a dissatisfying experience with a specific service category. Samples are drawn from university employees, including all types of occupational classifications (e.g., administrative, union), excluding faculty. University employees include a large number of faculty members who tend to have higher levels of education and income. This particular group may bias the study results, thus, faculty members were excluded from the sample. Two sets of self-instructed paper and pencil questionnaires were developed with minor wording adjustments depending on the service context. The survey packets were distributed via campus mail to 4,000 employees (2,000 for each service context) randomly chosen from the directory of entire 12,200 employees. A total of 240 surveys were returned (112 auto repair and 128 medical services).

3.2 Sample descriptions

Across the two service contexts, the majority of respondents are female (74 %) and the average age is 45 years. The majority of respondents (70 %) are married, 22 % of respondents are single, and 8.4 % are divorced or widowed. For the education level, 39 % of respondents have a college degree; 21 % have some college education; 16 % have a high school degree; and 24 % have a graduate degree. For income, 18 % of respondents indicate an income level of $35,000 or lower; 44 % report $36,000–75,000; and 38 % are above the $75,000 income level. Overall, demographic characteristics are fairly consistent across the two service contexts. Table 1 summarizes the demographic characteristics of the sample by the service context.

Table 1 Sample characteristics
Full size table

3.3 Classification of dissatisfied customers

To determine the two types of non-complainers (patience and neglect) and the two types of complainers (recovery satisfied and recovery dissatisfied), categorical variables (classification questions) and other measures are included.

3.3.1 Complaint behavior

A single item, dichotomous (yes/no) variable is used. The respondents were asked “did you complain to the service provider directly at that time?”

3.3.2 Neglect and patience

The respondents who answered “No” for the complaint behavior were asked to further indicate one of two responses that best described their reaction: “I did not complain because I did not care about the company and it was not worth taking any actions” for neglect or “I did not complain because I wanted to give the service provider/company another chance” for patience. These two categorical descriptions were based on the works of Rusbult et al. (1982), Ping (1993), and Dewitt and Brady (2003).

3.3.3 Recovery perceptions

The respondents who answered “Yes” for complaint behavior were asked to rate the service recovery that they had received. A six-item service recovery justice scale was adapted from Blodgett et al. (1997) and (Smith et al. 1999) and they were measured by a seven-point scale (1 = strongly disagree, 7 = strongly agree) (α = .861).

Based on the answers to the questions above, respondents were categorized into four groups: (1) neglect, (2) patience, (3) recovery satisfied, and (4) recovery dissatisfied. “Neglect” customers are the ones who did not complain because they did not care about the company and it was not worth taking any action (n = 45). “Patience” customers are the ones who did not complain because they wanted to give the company another chance (n = 45). For the complainers who received a service recovery, a median split (Median = 4.0) was performed on the recovery perception variable to categorize complainers into two groups: recovery satisfied and recovery dissatisfied. As expected, mean recovery satisfaction was significantly higher for recovery satisfied group (M = 5.31) as compared to recovery dissatisfied group (M = 2.58) (t = 6.46, p < .000). “Recovery satisfied” are the customers who complained and had a satisfying service recovery (n = 65) while “recovery dissatisfied” are the customers who complained and had a dissatisfying service recovery (n = 75). Out of 240, ten participants who failed to answer the follow up questions were excluded, resulting in 230 samples for further analyses (see Table 2).

Table 2 Participants by service context
Full size table

3.4 Measures

3.4.1 Attitude toward complaining

Respondents’ general complaining tendency was measured via a seven-item scale (e.g., “It sometimes feels good to get my dissatisfaction and frustration with the product off my chest by complaining”) from Singh (1990) and they were measured by a seven-point scale (1 = strongly disagree, 7 = strongly agree) (α = .756). The scales used in the study can be found in the “Appendix”.

3.4.2 Criticality of service failure

Three items (mild-severe; not important- very important; not at all dissatisfied-completely dissatisfied; seven-point scale) adapted from Hess et al. (2003) were used to assess the criticality of the service failure (α = .837).

3.4.3 Emotional bonding

A two item, seven-point scale was adapted from Mattila (2006). Respondents were asked to indicate the level of their emotional bonding with the service provider prior to the service failure incident. The scale ran as follows: “my level of emotional attachment to this auto shop/medical service facility was” (1 = much lower than average, 7 = much higher than average) and “the strength of my commitment to my relationship with this auto shop/medical service facility was” (1 = very low, 7 = very high), α = .942.

3.4.4 Return intentions

A two item, seven-point scale adapted from Blodgett et al. (1997) was used to measure return intentions, “how likely is it that you would return to that auto shop/medical service facility in the future?” (1 = very unlikely, 7 = very likely) and “because of what happened, I would never go to this auto shop/medical service facility again” (1 = strongly disagree, 7 = strongly agree), α = .910.

Prior to testing hypotheses, confirmatory factor analysis assessed the study variables’ psychometric properties by estimating a measurement model of the variables (see Table 3). Results indicate that the model offers a good fit to the data (χ 2 = 154.22, df = 70, CFI = .945, NFI = .905, and RMSEA = .070. Construct reliability estimates are high (.76–.94). All scales demonstrate adequate convergent validity and discriminant validity: the AVE for each variable is above .50 except for attitude toward complaining, and exceeds the shared variance (Fornell and Larcker 1981).

Table 3 Descriptive statistics and confirmatory factor analysis results
Full size table

4 Results

4.1 Dissatisfaction responses

Multinomial logistic regression was employed to test H1–H3. In the model, attitude toward complaining, criticality of the service failure, emotional bond, and service context were entered by treating complaint as the reference category. The model was significant (χ 2 = 52.268, p < .001) and the Pearson Goodness-of-Fit test indicated that the model fit the data adequately (p = .715). Negelkerke R square was .237 and the classification table indicated that 63.4 % of the cases for the dependent variable were correctly predicted by the model. Based on the Likelihood ratio tests of the overall effects, attitude toward complaining, criticality of the service failure, and emotional bond were found significant (p < .05). However, there was no significant difference by service context (p = .123). The results including parameter estimates, Wald statistics, p value, and odds ratio or Exp(b) are shown in Table 4.

Table 4 Multinomial logistic regression analysis results
Full size table

Attitude toward complaining and emotional bonding were significant in distinguishing neglect from complaint (p < .05) but they were not significant in distinguishing patience from complaint. Criticality of the service failure, on the other hand, was significant in distinguishing patience from complaint but it was not significant in distinguishing neglect from complaint (p < .001). This means that customers with a higher attitude toward complaining and emotional bonding were less likely to choose neglect response over complaint. For each unit increase in attitude toward complaining, the odds of choosing neglect response decreased by 38.7 % (.613–1.0 = −.387). Similarly, for each unit increase in emotional bonding, the odds of choosing neglect response decreased by 37.9 % (.621–1.0 = −.379). Also customers who had experienced a more critical service failure were less likely to choose patience response over complaint (p < .001). For each unit increase in criticality of service failure, the odds of choosing patience response decreased by 43.8 % (.562–1.0 = −.438). These results provide partial support for H1–H3.

In order to gain further insight regarding the difference in emotional bonding between patience and neglect, an independent t test was performed. There was a significant difference in emotional bonding between patience (M = 4.478) and neglect (M = 3.256) (t = 3.425, p < .01). This finding provides additional support for the idea that neglect reflects emotional withdrawal whereas patience reflects emotional attachment.

In summary, the likelihood that a customer will react to a service failure via neglect, patience, or complaint is related to his/her attitude toward complaining, perception of the service failure criticality, and emotional bonding. Specifically, attitude toward complaining and emotional bonding were negatively associated with the likelihood of neglect while criticality of the service failure was negatively associated with the likelihood of patience, compared to complaint.

4.2 Return intentions

Return intentions were examined via four customer groups and the service context. Since criticality of the service failure is known to influence customers’ post-failure responses (Webster and Sundaram 1998; Weun et al. 2004), it was included as a control variable. The ANCOVA results (Table 5) indicated a significant main effect of the service context (F = 46.843, p < .001) and the overall return intentions for medical services were higher (M = 5.381) than for auto repair services (M = 3.604). These findings provide support for H6. There was also a significant main effect for the customer group (F = 10.223, p < .001) but the interaction between the customer group and the service context was not significant (p = .363).

Table 5 Return intentions by customer group and service context
Full size table

Next, planned contrasts were performed to test hypotheses 4 and 5. The results show that “patience” response customers (M = 5.190) have higher return intentions than “neglect” response customers (M = 3.664) and the mean difference was statistically significant (difference = 1.527, p < .001). Also, they have higher return intentions than “recovery dissatisfied” customers (M = 3.839) and the mean difference was statistically significant as well (difference = 1.351, p < .01). Thus, hypotheses 4 and 5 are supported. Further examinations reveal that there were no statistical differences between the “patience” and “recovery satisfied” groups (difference = .087, p = .819) and the “neglect” and “recovery dissatisfied” groups (difference = .176, p = .629). See Fig. 2 for means by customer group and service context.

Fig. 2
figure 2

Return intentions by service context and customer group

Full size image

5 Discussion

Understanding a customer’s reaction to a service failure is an important but challenging issue for the service firms, especially given that some individuals exhibit behavioral responses, whereas others choose not to react in relatively similar dissatisfying service experiences (Singh 1988). By differentiating patience and neglect, this study offers a richer understanding of the customer dissatisfaction responses that provides a useful foundation for further research on non-complainers.

This study identifies that some non-complainers might be willing to give the service provider another chance by believing that the situation will improve in the future, while others might simply not care about the company. Attitude toward complaining, criticality of the service failure, and emotional bonding were useful predictors for differentiating complainers and non-complainers. In other words, attitude toward complaining, emotional bonding, and critical service failure lead customers to voice their dissatisfaction instead of being silent. Particularly, attitude toward complaining and emotional bonding reduce the likelihood of the neglect option compared to complaint. On the other hand, criticality of the service failure reduces the likelihood of the patience option compared to complaint. These findings suggest that when customers have a positive attitude and high emotional bonding with companies, they are less likely to be indifferent about the service failure but it leads them to voice their dissatisfaction on the spot. However, when the service failure is critical, customers are less likely to patiently suffer through it but voice their dissatisfaction instead. In a direct comparison of patience and neglect in terms of emotional bonding, patience response customers showed higher emotional bonding than neglect response customers. This provides additional distinction between the two non-complainers’ dissatisfaction responses, patience as emotional attachment and neglect as emotional withdrawal.

According to the results, the patience response customers showed higher return intentions than the neglect response customers. This result can be understood in that the patience customers endure the negative situation, hoping the situation will improve in the future and to continue their future relationship with the company. Whereas, the neglect customers give up on complaining to the service provider believing that it would not make much difference, and thus they are avoiding current and future interactions with the company. The study also indicates that the patience response customers’ return intentions are as high as those of the satisfying service recovery customers and higher than those of customers with neglect responses and dissatisfying service recoveries. These results are particularly noteworthy because when complainers and non-complainers are collectively compared, there is no significant difference in return intentions between them (t = 1.389, p = .159). Yet, when non-complainers were divided into two groups (neglect and patience) and complainers were also further separated into two groups (satisfied with recovery and dissatisfied with recovery), a clear pattern emerged.

The results also offer some insights into the mixed results regarding the non-complainers’ return intentions in past research. This distinction between patience and neglect may suggest that the high return intentions of non-complainers in the past research (e.g., Voorhees et al. 2006) might have reflected a patience response. On the contrary, the low return intent rates in other studies (e.g., TARP 1999) might have been driven by a neglect response. Thus, this research emphasizes differentiating patience and neglect responses when studying non-complainers.

Finally, return intent in the industry context offers useful insight to researchers and practitioners. Overall, return intentions for medical services were higher than for auto repair services. The findings reflect that the customer dissatisfaction responses in this monopolistic market differ from those in competitive markets (Roos et al. 2004; Singh 1990; Tronvoll 2007). Medical services are relatively less competitive than auto repair services, and thus customers have fewer alternatives. Moreover, insurance policies restrict customers’ ability to switch medical service providers.

6 Managerial implications

The study results demonstrate that non-complainers with a patience response are more likely to stay with the company, even after a dissatisfying experience without a service recovery, because they want to give the service provider another chance. It is somewhat comforting to know that even though a majority of dissatisfied customers fail to communicate their dissatisfaction to a service provider, some customers are likely to return as long as there is an emotional bond between the customers and companies. However, without an emotional bond, non-complainers are described as neglect and their return intentions are reduced. Therefore, these findings stress that service providers cultivate a relationship with customers on an emotional level.

For the relationship building, employees are often in a role to detect, prevent, and remedy dissatisfaction, thus, effective complaint management is dependent on employee characteristics (e.g., pro-social behavior, positive affect, knowledgeable) and organizational support factors (e.g., empowerment, customer-oriented culture, employee incentives, openness to hearing complaints) (McAlister and Erffmeyer 2003). Enhancing internal marketing quality through organizational and supervisory efforts to support employees is suggested to reduce customer complaints and strengthen the relationship between customers and firms (Bell et al. 2004).

7 Limitations and future research

There were several limitations to this study. First, the samples for this study were drawn from a university staff pool and the response rate was low. In order to assess the non-response bias, early and late respondents were compared for their demographics and no difference was found. However, to the extent that these respondents may be different from the consumers in other professions, generalizability of findings is affected. Alternatively, a larger sample of random sampling of households would be desirable to increase generalizability by reflecting the more diverse backgrounds of respondents.

Second, the findings of this study suggest that the emotional bonding level is higher for patience than neglect and that the lack of emotional bonding leads customers to avoid complaint by choosing neglect. There is, however, some evidence to suggest that the emotionally bonded customers might feel betrayed when a service failure occurs and engage in retaliatory actions (Grégoire et al. 2009). Phau and Baird (2008) suggest that both complainers and non-complainers engage in some form of retaliatory responses. Further research is needed to clarify the buffering and magnifying effects of emotional bonding on customers’ dissatisfaction responses. In addition, emotional bonding and return intentions constructs were assessed via only two items. Using more items to assess the nature and strength of emotional bonding and return intentions would be desirable for future examinations.

Third, in this study, dissatisfied customers’ responses were categorized into four distinct and mutually exclusive types by focusing on the immediate reactions to the service provider during the service failure incident. However, after the incident, customers may engage in other complaining behaviors, such as negative word-of-mouth and third-party actions. Further studies on the aftermath behaviors can provide useful insights into the multiplex and sequential nature of complaining behaviors. It is expected that the neglect response is more likely to be related to subsequent complaining behaviors such as negative word-of-mouth and third-party actions than the patience response. Another limitation of the four-category scheme is that an organization initiated recovery for non-complainers is not included. Customers who did not complain but experienced organization-initiated recoveries exhibit enhanced perceptions of justice (Voorhees et al. 2006). Therefore, this type of non-complainers may have higher return intentions than the patience and neglect response customers. Future studies are encouraged to extend return intentions of non-complainers with additional categories.

Finally, two service contexts (auto repair and medical) were used in the current study. Future research spanning various segments of the service industry is warranted to see if this pattern holds for the other types of services as well and which industry has more patience or neglect responses among non-complainers. In addition, more and more service encounters are infused with technology which certainly changes the nature of the service encounters with them becoming less interpersonal. This trend may suggest more apathy among customers toward service firms. Since neglect is considered to be a beginning of the customer alienation (Oliver 1997), future research is suggested to study the neglect and patience responses that involve technology based service encounters.