INCREASING CUSTOMER SERVICE BEHAVIORS USING MANAGER-DELIVERED TASK CLARIFICATION AND SOCIAL PRAISE

Abstract

This project assessed an intervention to improve employee customer service behaviors (correct greetings and closing behaviors). A combination of task clarification and manager-delivered social praise resulted in increased correct greeting from 11.5% to 66% and correct closing from 8% to 70%. The effect was maintained at a 48-week follow-up for employees who were present during the initial study period, but not for more recently hired employees. The results suggest that task clarification combined with manager-delivered social praise is an effective way to improve employee customer service behaviors.

The United States economy has changed from a product to a service economy, with 55% of economic activity accounted for by the service industry (U.S. Census Bureau, 2006). Almost three quarters of all customer purchases are made by repeat customers (Barlow & Maul, 2000). In addition, only 14% of customers who switch providers do so because they are unhappy with the product; customers generally switch because they are dissatisfied with the service they receive. Therefore, offering excellent customer service is one way organizations survive in an increasingly competitive economy. Retaining customers is important because the cost of gaining a new customer is nearly five times that of keeping an existing one (Barlow & Maul).

A number of studies have successfully improved customer service behavior. For example, Brown, Malott, Dillon, and Keeps (1980) targeted customer service behaviors by implementing two interventions—task clarification followed by feedback—and found that task clarification produced only a slight increase in the customer service behaviors whereas feedback produced an 87.7 percentage-point increase. Crowell, Anderson, Abel, and Sergio (1988) targeted customer service behaviors among bank tellers using task clarification, graphic and verbal performance feedback, and praise. Task clarification produced a 12% increase, feedback produced an additional 6% increase, and praise resulted in an additional 7% increase.

Previous research suggests that task clarification produces small effects in some cases and little or no behavior change in other cases. These differences may have occurred because previous studies did not take a functional approach to behavior change (i.e., the studies did not evaluate the potential cause of poor employee performance). Therefore, the interventions may have focused on improving behavior that was not the cause of poor employee customer service. By contrast, a functional approach would address the causes of behavioral deficiencies.

Another limitation of studies of organizational behavior management (OBM) is that researchers or practitioners, as opposed to the employees and managers themselves, have implemented described treatments. A recent literature review by Sigurdsson and Austin (2006) suggested that studies that involved the targeted organizational personnel in the development and implementation of data collection and consequence systems were more likely to report behavioral maintenance without direct experimenter intervention.

The purpose of the current study was to extend the literature on improving customer service behavior in the following ways. First, we conducted a functional assessment of the workplace environment to determine the potential causes of poor customer service. Second, we used the results of the functional assessment to develop an intervention that consisted of task clarification and social praise, which we trained the workplace manager to implement. Finally, we conducted follow-up at 48 weeks to determine the extent to which the effects of the intervention were maintained and generalized to new, untrained employees.

METHOD

Participants and Setting

Participants included 12 (both full- and part-time) workers (19 to 70 years old; tenure range, 0.5 to 22 months) at a grocery store located in the midwestern United States. During follow-up, 5 of the original 12 employees in addition to 5 new employees were observed. The store sold groceries and had a delicatessen, a coffee counter, a gift corner, and a dining area.

Dependent Variables

The dependent variables were correct greetings and correct closings, which were converted to a percentage after dividing the number of correct greetings or correct closings by the total number of observed greetings or closings. Each data point represented at least 10 opportunities for the employee to engage in a greeting or closing. A correct greeting was defined as the employee engaging in eye contact (employee looked at the face of the customer), smiling (the corners of the employee's mouth turned up), and greeting (the employee said, “Good morning,” “How are you today?” or “Welcome to [name of the store]”). A correct closing was defined as the employee engaging in eye contact, smiling, verbal closing (the employee said, “Good bye” or “Have a nice day”), and thanking the customer. Employees had to engage in all of the components of the greeting or closing behavior for the observer to score that behavior as correct.

The observer sampled the behavior of a subset of employees on each observation day. An observer who only sat near the front door and the coffee counter in an area in which customers dined, collected data on a data sheet at various arbitrarily selected times of day and days of the week until he or she recorded at least 10 customer interactions each for greeting and closing. Employees were not aware that they were being observed.

Interobserver Agreement

Occurrence and nonoccurrence interobserver agreement was calculated for 53% of the sessions, and mean agreements were 99% and 98%, respectively. The second observer was the manager, and he was trained on the operational definitions of the target behaviors before collecting any interobserver agreement data.

Functional Assessment

An informal functional assessment was conducted using the Performance Diagnostic Checklist (PDC; Austin, Carr, & Agnew, 1999). The functional assessment was based on questions in four areas: antecedents and information (e.g., Has the employee received adequate instruction?), equipment and processes (e.g., If equipment is required is it reliable?), knowledge and skills (e.g., Can the employees physically demonstrate the task?), and consequences (e.g., Are there consequences delivered contingent on the task?). The PDC results revealed that there were no written descriptions and job or task aids to inform the employees of the expected behaviors or when to engage in them (i.e., antecedents and information); in addition, consequences provided by managers were not delivered frequently after proper greetings and closings, and managers did not appear to deliver feedback or regularly monitor customer service performance (i.e., consequences). The function-based, manager-implemented intervention package consisted of task clarification to inform employees of expected behaviors and social praise following correct greetings and closings.

Experimental Design

A multiple baseline design across behaviors (Petursdottir, Carr, Lechago, & Almason, 2008) was employed to evaluate the effects of task clarification and social praise on greeting and closing behaviors for the group of employee participants.

Procedure

Baseline

Baseline was observation only. No changes were made to employee behavior. The only change made to the manager's behavior was that he collected interobserver agreement data.

Task clarification and social praise

The manager delivered task clarification for greetings and for closings using a script for each set of behaviors. The manager provided task clarification (which took approximately 7 min) to each employee individually at a table in the dining room. Task clarification occurred over a 3- or 2-day time frame for greeting and closing, respectively. The experimenter trained the manager to recognize a correct greeting or closing and to approach the employee and give social praise following a correct greeting or closing by saying, “Great job on your greeting,” or “That was great customer service.”

Follow-up

The first author collected data on correct greetings and closings 48 weeks after the task clarification and social praise phase ended. The observer collected data separately for employees who were present during the initial study period and employees who were hired after the study period had ended to determine if there was a difference between these groups.

Integrity of the Independent Variable

After the manager trained each employee, he had the employee initial and date the script to ensure that the training had been delivered to everyone. All employees signed the greeting script. Seventy-nine percent of employees signed the closing script. Instances of manager-delivered social praise were observed during 10 different sessions.

RESULTS AND DISCUSSION

Figure 1 (top) shows the percentage of correct greetings during baseline and task clarification and social praise conditions. During baseline the mean percentage of correct greetings was 11.5% (range, 0% to 40%; SD  =  11.4%). During the task clarification and social praise treatment, the mean percentage of correct greetings increased to 66% (range, 27% to 82%; SD  =  15%). At the 48-week follow-up, mean percentage of correct greetings was 70% (range, 60% to 80%; SD  =  10%) for employees who were exposed to the treatment during the initial study period and was 46.7% (range, 40% to 50%; SD  =  5.8%) for the new employees.

Figure 1.

The percentages of correct greeting (top) and closing (bottom) behaviors.

Figure 1 (bottom) displays the percentage of correct closings during baseline and task clarification and social praise conditions. Mean percentage of correct closings was 8% (range, 0% to 30%; SD  =  9.8%) during baseline. After task clarification and social praise were initiated, mean percentage of correct closings was 70% (range, 60% to 90%; SD  =  9%). At the 48-week follow-up, mean percentage of correct closings was 76.7% (range, 60% to 90%; SD  =  15.3%) for employees who had been exposed to the treatment and was 40% (range, 30% to 50%; SD  =  10%) for the new employees.

The experimenter gave the manager the closing and greeting scripts with which he could train new employees and the data-collection sheet in electronic format so that he could continue collecting data and delivering social praise. Although this was an attempt to set the occasion for maintenance, our postintervention checks revealed that the manager did not use the scripts for training new employees and did not continue to deliver social praise. The manager reported that he did not prefer to deliver social praise, which may be why it was not maintained. This low acceptability of social praise is a limitation of the techniques described in this study.

The 48-week follow-up data indicated that behavior change was maintained for employees exposed to the treatment but not for new employees. Interestingly, new employees engaged in higher levels of correct greetings and closing than the original employees did during baseline. The manger reported that he did not use the scripts in training new employees or give them social praise, but he may have communicated to them that they were expected to engage in the greeting and closing behaviors or given social praise at some point. Also, new employees may have observed their coworkers engaging in these behaviors and assumed that they were to engage in them as well.

Social praise was difficult to implement for the manager because it required him to be in the vicinity of the customer–employee interaction for the duration of the interaction to fully observe the behaviors being emitted. Even if the manager was in the vicinity of the interaction, he often had to leave for various reasons, and as a result, he missed the praise opportunity. In addition, the manager did not have the opportunity to give praise between occurrences if there were consecutive customer interactions. Future studies could seek to create and implement a system in which praise could be given more frequently and easily and in ways more likely to be accepted by managers.

A limitation of the current study is that the intervention consisted of two components, task clarification and social praise. It is not clear if both components were required to produce behavior change. Future studies should attempt a component analysis of the intervention. Further, this study does not validate the PDC as a functional assessment tool. It is possible that interventions based on deficiencies that were not identified (e.g., equipment and processes) also would have resulted in behavior change. Future studies should compare interventions based on the outcomes of the PDC to arbitrarily selected interventions or to interventions contraindicated by the PDC to better validate the PDC as a functional assessment tool. Finally the current study incorporated only two parts of a multiple baseline design, and thus only limited internal validity can be assessed.

In summary, the present data suggest that task clarification combined with social praise substantially increased employee customer service behaviors and corroborate the findings of other studies (e.g., Crowell et al., 1988; Komaki, Barwick, & Scott, 1978) when the intervention was applied to greeting behaviors.