The Relationship Between Customer Complaints to the Firm and Subsequent Exit Behavior

ABSTRACT - This paper investigates the influence of customers' complaints to a service firm with their subsequent decision to exit from a relationship with the firm. The context is a longitudinal study of cellular communications customers' behavior. The study shows that overall satisfaction with a firm and customer complaint behavior are related to exit behavior. Furthermore, companies should strive to "get it right the first time" because the firm may not always be able to appease a complaining customer.

Citation:

Ruth N. Bolton and Tina M. Bronkhorst (1995) ,"The Relationship Between Customer Complaints to the Firm and Subsequent Exit Behavior", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 94-100.

Advances in Consumer Research Volume 22, 1995      Pages 94-100

THE RELATIONSHIP BETWEEN CUSTOMER COMPLAINTS TO THE FIRM AND SUBSEQUENT EXIT BEHAVIOR

Ruth N. Bolton, GTE Laboratories Incorporated

Tina M. Bronkhorst, GTE Laboratories Incorporated

ABSTRACT -

This paper investigates the influence of customers' complaints to a service firm with their subsequent decision to exit from a relationship with the firm. The context is a longitudinal study of cellular communications customers' behavior. The study shows that overall satisfaction with a firm and customer complaint behavior are related to exit behavior. Furthermore, companies should strive to "get it right the first time" because the firm may not always be able to appease a complaining customer.

As part of a defensive marketing strategy, complaint management is designed to minimize turnover (Fornell and Wernerfelt 1987). Complaint management entails finding and correcting systemic causes of dissatisfaction rather than simply placating unhappy customers. Although there is a vast literature on customer complaining behavior, surprisingly few studies have investigated the relationship between customer complaints and turnover. This paper investigates the influence of customers' complaints on their subsequent decision to exit from a relationship with a firm.

BACKGROUND

Customer complaining behavior (CCB) has been defined as "a multiple set of behavioral and nonBbehavioral responses triggered by a dissatisfactory purchase episode" (Singh 1988, p. 94). A substantial body of research has focused on the nature and classification of these responses. Hirschman (1970) classified customers' responses to dissatisfaction as exit, voice or loyalty. In buyerBseller relationships, a customer may exit from a relationship with a firm by switching brands or service providers, or by reducing consumption levels. He/she may voice complaints to friends, thirdBparty organizations (e.g., Better Business Bureau) or sellers (i.e., retailers or manufacturers). Lastly, a customer may choose not to act, thereby remaining loyal. In subsequent studies, market researchers have extended this classification scheme. For example, they have developed typologies that differentiate between public and private voice (Day and Landon 1977) and that discriminate among the objects toward which the response was directed (Singh 1988). A recent experiment provides empirical support for the validity and usefulness of Hirschman's taxonomy (Maute and Forrester 1993).

There is a vastCalbeit fragmentedCliterature on voice responses (e.g., Strahle and Day 1985). In fact, Andreason (1988) estimates that there are over 500 studies on predictors of CCB responsesCthat is, studies investigating the dual questions of which customers complain and why. Several empirical studies have shown that customer dis/satisfaction is related to customer complaining behavior (Bearden and Teel 1983; Oliver 1980; Olshavsky and Miller 1972). Mediating variables include causal attributions about the service failure, attitudes and demographics, the costs and benefits of complaining frequency of interactions, the product/service context, and situational factors (Andreason 1977; Day and Landon 1977; Folkes 1984a; 1984b; Fornell and Didow 1980).

This paper investigates the relationship between voice and exit responses, focusing on complaints to the seller. Although voice and exit responses are frequently considered to be substitute responses to dissatisfaction, they may actually be complements. However, few studies have examined their relationship. In a notable exception, Solnick and Hemenway (1992) found that customers who register medical complaints are four and a half times more likely to leave their Health Maintenance Organization than customers who do not complain. Furthermore, in an early study, Gilly and Gelb (1982) established that customers who were more satisfied with the firm's response to their complaints were more likely to repurchase, and tended to have slightly higher reBpurchase levels.

PROPOSITIONS

This study focuses on three facets of the relationship between complaints and exit behavior. First, it examines the relationship between the incidence of customer complaints and subsequent exit behavior. Then, it investigates how the nature of customer complaints mediates this relationship. Last, it looks at the potentially ameliorating effect of sellers' attempts at redress.

Complaints and Exit Behavior

By definition, CCB arises from product/service dissatisfaction. Prior studies have found that higher levels of dissatisfaction with the firm are associated with increased brand switching behavior (LaBarbera and Mazursky 1983) and exit intentions (Maute and Forrester 1993). However, there is little research relating customer dis/satisfaction to actual behavioral measures. Hence, our first proposition concerns the relationship between satisfaction and exit behavior.

P1: Customers who have lower levels of overall satisfaction with the firm are more likely to exit.

As discussed earlier, there is a paucity of studies about the relationship between voice and exit responses (Andreason 1985). However, Solnick and Hemenway (1992) found that patients who register medical complaints are four and a half times more likely to voluntarily exit from a Health Maintenance Organization.

P2a: Customers who complain are more likely to exit than customers who do not complain.

Although it is seldom discussed, an alternative form of exit behavior is a reduction in consumption levels. This behavior is particularly likely in monopoly conditions.

P2b: Customers who complain will show a decrease in consumption levels relative to customers who did not complain.

The Nature of Customer Complaints

In a study of credit card purchases by major oil company customers, Gilly and Gelb (1982) found that complainants are more likely to be satisfied by the company's response to monetary problems, and they are more satisfied if the percent received is higher. Thus, we might expect that they are less likely to exit.

P3: Customers who complain about monetary loss (e.g., billing errors) are (a) more likely to be satisfied with the firm's response and (b) less likely to exit than customers who complain about other problems.

Prior research has shown that customers' attributions about the stability, locus and controllability of product failure are useful in predicting expectancy reactions, marketplace equity reactions and anger reactions (Folkes 1984b). These effects should influence customers' satisfaction with sellers' efforts at redress. Customers should be less satisfied with sellers' redress of problems that are firmBrelated (i.e., the locus of the cause is within the firm) and when the firm has control over the reason for the product failure. Hence, they should be more likely to exit.

P4: Customers who complain about problems that are firm related and within the firm's control are: (a) less likely to be satisfied with the firm's response and (b) more likely to exit than customers who complain about other problems.

The Effect of Sellers' Attempts to Redress Complaints

From the standpoint of the firm, attempts at redress are designed to reduce the likelihood that a customer will exit. A variety of studies have investigated how the sellers' responsiveness to complaints affects complainants' satisfaction, perceptions of fairness, and so forth (e.g., e.g., Goodwin and Ross 1992; Smart and Martin 1992). There is less evidence about how sellers' redress efforts affect exit behavior. However, Gilly and Gelb (1982) found that the higher the degree of satisfaction with organization complaint response, the greater the likelihood of brand reBpurchase. This finding may extend to other products.

P5: The higher the degree of satisfaction with the firm's response to a complaint, the greater the likelihood that the customer does not exit.

STUDY DESIGN

This study investigates customer voice and exit behavior in the rapidly growing cellular communications industry. Although customers have a monthBtoBmonth subscription relationship with their cellular provider, over 30% of the typical cellular providers' customers exitCeither switching providers or discontinuing cellular usageCeach year (Steuernagel 1992). Hence, the cellular industry provides a particularly appropriate context for an investigation of voice and exit responses.

In the customer complaint literature, studies have typically measured complaint intentions and reBpurchase intentions. In contrast, this study measures actual complaint behavior and subsequent exit behavior by tracking customers who subscribe to cellular service from a large firm over an 18 month period. During January through March 1992, a probability sample of 3361 customers (who had subscribed for three months or more) were interviewed using a standard customer satisfaction survey questionnaire. (This survey is administered by a commercial market research supplier each quarter.) Selected items from the questionnaire are displayed in Exhibit 1. These data were merged with company billing records that capture customers' subsequent consumption behavior and exit behavior for the period January 1992 through August 1993.

Customer Satisfaction and Complaints. As shown in Exhibit 1, customers were asked to rate their overall satisfaction with the firm's service, to report their perceptions of any calls they made to customer service centers in the past three months, and to provide demographic information. Customer complaints were identified from a survey question about the purpose of calls to customer service centers (Q9h) administered prior to any decision to stay loyal or exit from the firm. "A question regarding your bill," "a problem with your cellular service," "a problem with your phone equipment," and "a problem with roaming" were categorized as complaints. Since billing questions typically involve queries about the accuracy/appropriateness of charges, it was assumed that billing questions were problems/complaints.

Customer Exit Behavior. Between January 1992 and August 1993, the company tracked these 3361 customers to determine whether (and when) they subsequently cancelled service. In addition, each customer's monthly billing records measured consumption levels by average minutes of use per month.

RESULTS

Complaints and Exit Behavior

To investigate the first two propositions, data from two groups were compared: customers who exited and customers who did not (i.e., customers who remained "loyal.") The relevant data are shown in Tables 1 and 2.

P1: Customers who have lower levels of overall satisfaction with the firm are more likely to exit.

To examine P1, the distributions of responses to the overall satisfaction question (Q1) for the two groups were compared using a chiBsquare test. The results indicate that we should reject the null hypothesis that there is no difference between the groups (c²=33.03, p<.001). For example, 83.1% of customers who were "very satisfied" subsequently remained loyal; whereas, 37.0% of customers that were "very dissatisfied" subsequently exited.

P2a: Customers who complain are more likely to exit than customers who do not complain.

P2b: Customers who complain will show a decrease in consumption levels relative to customers who did not complain.

The two groups' exit behavior were compared with a chi-square test. As shown in Table 2a, the results indicate that there is a significant difference between the groups (c²=7.16, p<.01). Customers who complained were more likely to have subsequently exited than customers who did not complain. This result supports P2a. The differences in the two groups' change in consumption levels were compared in the following way. As shown in Exhibit 1, customers' survey responses indicated whether they had called customer service "in the past three months." Since the interviews took place in January through March 1992, the time period during which the complaint was registered ("the past three months") roughly corresponds to the November 1991 through January 1992. The time period after which the complaint was registered roughly corresponds to March through May 1992. Hence, each customers' average monthly minutes of use was computed for the periods roughly "during" the complaint was registered and "after" the complaint was registered. Then, the difference between these two consumption levels was calculated for each customer. Average consumption levels decreased for both complaining and nonBcomplaining customers due to seasonality. More importantly, the null hypothesis that the average change in the consumption levels of the two groups is equal could not be rejected (p=0.39). However, the direction of the effects is consistent with our proposition. As shown in Table 2b, complainants show an average decrease of 15 minutes/month whereas nonBcomplainants show a decrease of only 11 minutes/month. In conjunction with P2a, these results suggest that cellular customers' complaints are more likely to be followed by cancellation of service than by reduced consumption levels. This finding is not entirely surprising because customers can usually subscribe to service from another provider.

EXHIBIT 1

SURVEY AND BILLING ITEMS

TABLE 1

OVERALL SATISFACTION WITH FIRM AND EXIT BEHAVIOR

TABLE 2A

CUSTOMER COMPLAINING BEHAVIOR VERSUS EXIT BEHAVIOR

TABLE 2B

CUSTOMER COMPLAINING BEHAVIOR VERSUS MINUTES OF CELLULAR USE

TABLE 3A

BILLING COMPLAINTS VERSUS SATISFACTION WITH THE SERVICE REPRESENTATIVE

TABLE 3B

EXIT BEHAVIOR VERSUS BILLING COMPLAINTS

The Nature of Customer Complaints

Cellular customers may complain about three types of problems: billing, service and equipment problems. To investigate P3 and P4, we compare these different categories of complaints.

P3: Customers who complain about monetary loss (e.g., billing errors) are: (a) more likely to be satisfied with the firm's response and (b) less likely to exit than customers who complain about other problems.

This proposition is investigated by comparing the responses of customers who reported billing complaints to customers who reported other types of complaints. The null hypothesis that the two groups' satisfaction with the firm's handling of the complaint are equivalent is tested using their ratings of the customer service representative (Q9c). The data in Table 3a indicate that the null hypothesis cannot be rejected; there is no difference between the groups (c²=7.16, p<.15). In fact, 57.8% of respondents with a billing complaint were "very satisfied" with the service representative whereas 64.5% of respondents with other complaints were "very satisfied." This finding is inconsistent with P3a.

To examine P3b, the two groups' exit behavior is compared. As indicated in Table 3b, the null hypothesis that there is no difference between the groups cannot be rejected (c²=2.05, p<.15). However, customers who call with billing complaints tend to have slightly lower exit levels than customers who call about other problems. Considering the results from P3a and P3b together, customers seem to rate the firm's handling of billing problems as no better or worse than their handling of other problems. Hence, even though they typically receive adjustments from their bill, they are no more likely to cancel.

P4: Customers who complain about problems that can be attributed to the firm and are within the firm's control are: (a) less likely to be satisfied with the firm's response and (b) more likely to exit than customers who complain about other problems.

Since the cellular provider does not manufacture or sell equipment, customers are unlikely to attribute equipment problems to the firm, or to believe that these problems are within the control of the firm. However, service problems (e.g., dropped calls, static on the line) should be attributed to the firm and are within the firm's control. Hence, this hypothesis is investigated by comparing the responses of customers who reported service problems to customers who did not report service problems. As shown Table 4a, the null hypothesis that there are no differences between the groups cannot be rejected (c²=3.14, p= 0.534). Customers who complain about service problems have equivalent satisfaction levels to customers who complain about other problems.

TABLE 4A

SERVICE COMPLAINTS VERSUS SATISFACTION WITH THE SERVICE REPRESENTATIVE

TABLE 4B

EXIT BEHAVIOR VERSUS SERVICE COMPLAINTS

TABLE 5

SATISFACTION WITH SERVICE REPRESENTATIVE VERSUS EXIT BEHAVIOR

To investigate P4b, the two groups' exit behavior is compared. The null hypothesis is rejected (c²= 5.72, p<0.02). Customers who complain about service problems are more likely to exit than customers who complain about billing or equipment problems. This result is intuitively appealing because these problems concern the "core service" of the firm.

The Effects of Seller's Attempt to Redress Complaints

To investigate the effects of the firm's handling of complaints, we examine the customer's perception of the sales representative (Q9c).

P5: The higher the degree of satisfaction with the firm's response to a complaint, the greater the likelihood that the customer does not exit.

As shown in Table 5, the null hypothesis that satisfaction with complaint handling and subsequent exit behavior are independent cannot be rejected (c²=4.54, p=0.338). Satisfaction levels for the service representative do not differ significantly between customers who exit and customer who remain loyal. This result demonstrates that the service representative cannot entirely forestall exit behavior by customers who have made a complaint.

DISCUSSION

This study shows a significant relationship between customer's overall dis/satisfaction with service and subsequent exit behavior. Customers who are generally dissatisfied with service are more likely to exit. Many companies regularly survey their customers and measure satisfaction levels across different operating units or geographic areas. These firms can target marketing and quality improvement programs at customer groups that report lower satisfaction levels, thereby maintaining the existing customer base, and potentially increasing profits. These profit increases may be substantial. In the cellular industry, for example, it is estimated to cost $600 to acquire a new customer and only $20 to retain an existing one (Steuernagel 1992). Consequently, it is far more profitable to retain existing customers than "replace" them.

Customers who contact a service company to voice a complaint are more likely to exit. This finding highlights the complementary nature of voice and exit behavior. Many companies believe that appropriate complaint handling can help them "recover" from customer complaints and retain customers. However, this study suggests that redress efforts are not always effective. Since redressing complaints to the customer's satisfaction is difficult the firm should "Get it right the first time" (Albrecht and Bradford 1990, p.101).

Unlike prior research, this study considered two forms of exit behavior: severance of the relationship with the firm (i.e., cancelling service) and a reduction in consumption levels (i.e., decreased minutes of cellular service use). Although customer complaints were not associated with significantly greater decreases in consumption levels, the direction of the effect was consistent with P2b. This finding is sufficiently interesting to warrant further investigation. Since the date of the complaint was unavailable, this study measured consumption levels before and after the customer's retrospective report of the complaint. In future research, it would be useful to examine consumption levels before and after the time that the complaint was registered at the firm.

Customers exit behavior varies depending on the nature of the complaint. When a disruption in core service is the source of the complaint, customers become dissatisfied and more likely to cancel. Customer expectations of ancillary services seem to be viewed less critically. Customers who report billing complaints do have slightly lower levels of exit (P3b).

The finding that customers who complained about service problems are more likely to exit than customers who complained about other (e.g., equipment) problems provides support for an attributionBbased approach complaining behavior. Although this study does not measure attributions, customers are found to be more likely to exit due to problems that can be categorized as firmB related and within the firm's control. This result is consistent with studies of customer satisfaction and service quality (e.g., Bitner 1990). For example, Bolton and Drew (1994) found that satisfaction with a repair process depended on locus of responsibility for the problem (firm versus customer), and whether the problem was within the firm's control (i.e., nonBequipment problems).

A firm's response to a complaint may not make up for the fact that a customer has been dissatisfied enough to voice a problem. Our investigation of P3a, P4a, and P5 all demonstrate that the service representative cannot necessarily "recover" from a complaint. Tables 3a and 4a both show more than 70% of respondents reporting that they were either "somewhat" or "very satisfied" with the service representative. As shown in Table 3a, customers are slightly more dissatisfied with firm's response to complaints regarding monetary loss than other types of complaints. Either service representatives are not be able to adequately address billing complaints, or customers may expect that the billing process should be "seamless" and resent billing errors and difficulties understanding the bill.

Managers may be somewhat surprised to find that, even when customers report that they are "very satisfied" with the customer service representative, they may still choose to exit (P5). There are two possible reasons for this result. First, not all complaints can be resolved by customer service representatives. Customers may perceive that the representative was courteous, knowledgeable and helpful; yet ultimately, the firm may have been unable to redress the underlying problem. For example, certain intermittent but recurring problems (e.g., static on the line for a cellular communications provider) may be very difficult to resolve. Second, even if a complaint is satisfactorily resolved by the firmCpossibly including monetary compensation for the disruption in serviceCthe customer will not necessarily be satisfied with overall service. From the customer's standpoint, it would be preferable if there had not been a disruption in the first place. Again, there is support for this notion in the literature on customer satisfaction and service quality (Bolton and Drew 1992).

CONCLUDING REMARKS

Despite considerable research on customer complaining behavior, few studies have examined the relationship between customer complaint and exit behavior. Even fewer studies have examined actual behavior, rather than intentions. The results from this study suggest that voice and exit behaviors are complementary. The study shows that overall dis/satisfaction with a firm is related to exit behavior. For this cellular provider, efforts to redress customer complaints are not sufficient to forestall exit behavior. This finding is indicative of an inadequate complaint resolution processes, and may not necessarily generalize to all firms. For example, as earlier studies have shown, some industries are characterized by factors such as high switching costs that impede exit behavior in the short run. However, this study also shows that exit behavior is likely to follow a complaintCeven when the complaint is handled satisfactorily. Thus, even firms with highly rated complaint handling processes should be concerned about forestalling complaints.

In this study, complainants show an average decrease of 15 minutes/month, whereas nonBcomplainants show an 11 minute/month decrease. The difference is not statistically significant. However, in the cellular industry, there are frequently two providers in a market and switching costs are relatively low, so that customers may choose to exit rather than reduce consumption. Nevertheless, the relationship between complaint behavior and consumption levels is worthy of further study in other industries. Consumption levels are directly tied into a corporation's profits, and this "alternative" form of exit behavior could be a costly problem. Many service companies collect consumption data for billing purposes, enabling them to study this issue. Such efforts would be an important step in understanding the costs and benefits of complaint management.

In the customer satisfaction and service quality literature, researchers have begun to developed sophisticated multivariate models that relate these constructs to their antecedents (e.g., Bitner 1990; Bolton and Drew 1992). A similar effort is required in the customer complaining literature. The natural extension of the work reported in this paper is a multivariate model of the relationship between customer complaints and exit behavior, and their antecedents.

REFERENCES

Albrecht, Karl and Laurence J. Bradford (1990), The Service Advantage, Homewood, Illinois: Dow JonesBIrwin.

Andreason, Alan R. (1977), "A Taxonomy of Consumer Satisfaction /Dissatisfaction Measures," Journal of Consumer Affairs, 11 (Winter), p. 11B24.

Andreason, Alan R. (1988), "Consumer Complaints and Redress: What We Know and What We Don't Know," in The Frontier of Research in the Consumer Interest, E. Scott Maynes et al, eds., Columbia, MO: American Council on Consumer Interests, 675B722.

Bearden, William O. and Jesse E. Teel (1983), "Selected Determinants of Consumer Satisfaction and Complaint Reports," Journal of Marketing Research, 20 (February), 21B8.

Bitner, M. J. (1990). "Evaluating Service Encounters: The Effects of Physical Surrounding and Employee Responses," Journal of Marketing, 54 (2), 69B82.

Bolton, Ruth N. and James H. Drew (1994), "Linking Customer Satisfaction to Service Operations and Outcomes," Service Quality: New Directions in Theory and Practice, Roland T. Rust and Richard L. Oliver, eds., 173B200.

Bolton, Ruth N. and James H. Drew (1992), "Mitigating the Effect of Service Encounters," Marketing Letters, 3 (1), 57B70.

Day, Ralph L. (1980), "Research Perspective on Consumer Complaining Behavior," in Theoretical Developments in Marketing, C. W. Lamb and P. M. Dunne, eds. Chicago: American Marketing Association, 211B15.

Day, Ralph L. and E. Laird Landon, Jr. (1977), "Toward a Theory of Consumer Complaining Behavior," in Consumer and Industrial Buying Behavior, Arch G. Woodside, Jagdish Sheth and Peter Bennnet, eds., New York, NY: ElsevierBNorth Holland, 425B437.

Folkes, Valerie S. (1984a), "An Attributional Approach to Postpurchase Conflict Between Buyers and Sellers," in Advances in Consumer Research, 11, Thomas C. Kinnear, ed. Ann Arbor, MI: Association for Consumer Resarch, 500B503.

Folkes, Valerie S. (1984b), "Consumer Reactions to Product Failure: An Attributional Approach," Journal of Consumer Research, 10 (4), 398B409.

Fornell, Claes and Nicholas M. Didow (1980), "Economic Constrains on Consumer Complaining Behavior," in Advances in Consumer Research, 7, J. C. Olsen, ed., Miami, FL: Association for Consumer Research 318B23.

Fornell, Claes and Birger Wernerfelt (1987), "Defensive Marketing Strategy by Customer Complaint Management: A Theoretical Analysis," Journal of Marketing Research, 24 (November), 337B 46.

Gilly, Mary C. and Betsy D. Gelb (1982), "PostBPurchase Consumer Processes and the Complaining Consumer," Journal of Consumer Research, 9 (December), 323B328.

Goodwin, Cathy and Ivan Ross (1992), "Consumer Responses to Service Failures: Influence of Procedural and Interactional Fairness Perceptions," Journal of Business Research, 25, 149B 63.

Hirschman, Albert O. (1970), Exit, Voice and Loyalty. Cambridge, MA: Harvard University Press.

LaBarbera, Priscilla A. and David Mazursky (1983), "A Longitudinal Assessment of Consumer Satisfaction/Dissatisfaction: The Dynamic Aspect of the Cognitive Process," Journal of Marketing Research, 20 (November), 393B404.

Maute, Manfred F. and William R. Forrester, Jr. (1993), "The Structure and Determinants of Consumer Complaint Intentions and Behavior," Journal of Economic Psychology, 14, 219B47.

Oliver, Richard (1980), "A Cognitive Model of the Antecedents and Consequences of Satisfaction Decisions, Journal of Marketing Research, 17, 460B9.

Olshavsky, Richard and John A. Miller (1972), "Consumer Expectations, Product Performance and Perceived Product Quality," Journal of Marketing Research, 9, 19B21.

Singh, Jagdip (1988), "Consumer Complaint Intentions and Behavior: Definitional and Taxonomical Issues," Journal of Marketing, 52(January), 93B107.

Singh, Jagdip (1990), "Voice, Exit and Negative WordBofBMouth Behaviors: An Investigation Across Three Service Categories," Journal of the Academy of Marketing Science, 19 (Winter), 1B15.

Smart, Denise T. and Charles L. Martin (1992), "Manufacturer Responsiveness to Consumer Correspondence: An Empirical Investigation of Consumer Perceptions," Journal of Consumer Affairs, 26, 104B128.

Solnick, Sara J. and David Hemenway (1992), "Complaints and Disenrollment at a Health Maintenance Organization," Journal of Consumer Affairs, 26 (1), 90B103.

Steuernagel, Robert A. (1992), "Limits to Growth," Cellular Business (July), 26B29.

----------------------------------------