Marc Preston Bernhisel and Dr. Kristie Seawright, Business Management

“Mistakes are an unavoidable feature of all human endeavor and thus also of service delivery” (Boshoff, 1997, p.110). Regardless of the efforts and precautions to avoid breakdowns and errors in service delivery, failures are bound to occur. Although failures will always occur, firms can avoid the cost of losing customers through service recovery. Zemke and Bell (1990) defined service recovery as a “thought-out, planned process for returning aggrieved customers to a state of satisfaction with the firm after a service or product has failed to live up to expectations” (p. 43).

While many acknowledge the importance of service recovery, recommendations on implementation differ. A good service recovery design can improve customers’ perception of the quality of products or services already purchased, the organization’s competence, and the value of the organization’s other offerings (Zemke and Bell, 1990). The difficulty remains of what exactly constitutes good service recovery.

The purpose of this research is to determine the most effective design process in achieving service recovery success by empirically examining various service recovery designs within environments of varying levels of service criticality.

This experiment utilizes three research propositions: 1) Service recovery success differs according to the level of service recovery design, 2) Service recovery success differs according to service criticality, or the importance of the original service, and 3) Service recovery success is affected by the interaction between service recovery design (psychological and tangible activities) and the criticality of the service.

A controlled experimental design was used to verify the primary and interaction effects of recovery design and service criticality on service recovery success. The design consisted of four treatments of service recovery efforts over three levels of service criticality (clothing, television, and automobile). In Treatment 1: Recovery Failure, the service provider offers a negative psychological effort by denying responsibility, with belligerence. In Treatment 2: No Recovery, the service provider offers neither a psychological nor a tangible recovery effort. In Treatment 3: Standard Recovery, the service provider employs both psychological efforts through communication, and tangible efforts through compensation, or completion of the original service. In Treatment 4: Value-Added Recovery, the service provider offers both psychological efforts and tangible compensation, and then reaches beyond compensation through returning 15 percent (10 percent for automobile) of the purchase price to pay for the inconvenience.

I obtained respondents by establishing a table in the center of the university student center. By collecting the data in the student center, I avoided the potential bias of reaching students in limited fields of study. I placed a sign on the table, which advertised the gift of one full-size candy bar for completing a ten- minute survey. The respondents did not know the content or purpose of the survey, prior to completion. I gave each respondent one survey, assuring that each respondent completed only one survey, without repeating.

Each survey contained a front page that provided the instructions, assurance of the volunteer status of participation, and the necessary contact information. The instructions directed the student to the next page, which contained one of the three failure scenarios (clothing, electronics, and automobile). The respondents were then directed to turn the page. The last page contained one of the four recovery treatments, followed by four more questions to measure satisfaction and loyalty following the recovery scenario.

The study examined a sample consisting of 12 groups with 29-30 consumers per group. Sample size was n=358. A 4×3 factorial analysis of variance (ANOVA) was employed to examine the data. ANOVA was used to determine if differences exist on the main effects of 1) recovery treatment, and 2) criticality. Also, ANOVA was used to determine if differences exist in the interaction of recovery treatment and criticality. Statistically significant differences were found on both main constructs, as well as on the interaction at the p< .01 levels. Service recovery success differences were found to be statistically significant based on treatment effects. This experiment confirms Research Proposition by showing that service recovery success does differ according to the level of service recovery design. Recovery success differences were also found to be statistically significant based on service criticality. The experiment therefore confirms Research Proposition 2 by showing that service recovery success does differ according to service criticality. Finally, recovery success differences were found to be statistically significant based on the interaction between recovery treatment and service criticality. The experiment therefore confirms Research Proposition 3 by showing that service recovery success is affected by the interaction between service recovery design and the criticality of the service. This study provides a foundation for service providers to practice effective service recovery efforts. Several of the findings supply businesses with valuable information regarding the service recovery process. First, this study found that service recovery design is important and significant across all levels of criticality. The study examined the recovery design treatments as a whole and found a significant difference, and then analyzed the individual treatments within an environment of criticality and found a significant difference. Therefore, the process design of a service provider’s recovery efforts truly does matter to retain the satisfaction and loyalty of consumers who have experienced a failure. This study also found that providing value-added atonement makes a difference. The satisfaction and loyalty ratings of respondents were significantly higher for consumers who received an additional act of atonement beyond the standard correction of the original service. The analysis showed that the effect of value-added atonement is the same across all three levels of criticality. Therefore, service providers of any type of service may benefit from offering more than the standard remedy for a service failure. Also, if service providers currently maintain a policy to merely provide a correction of the original service, they should be aware of consumer expectations regarding value-added atonement.