engleski

An Experimental Study of Balking and Reneging in Waiting Lines

Waiting on line for service is a common experience. Customers often balk or renege when they believe that waiting until they are served will delay them for their next task. Balking and reneging can create unhappy customers who might switch to a competing service provider. In this paper, we study the impact of various types of information on balking and reneging behavior in queues, using a computerized experiment. Participants in the experiment were told that they were at a post office to mail a gift, but they also had an appointment for which they might be late if the wait was too long. However, they could balk and return &#8220 ; ; tomorrow&#8221 ; ; or, if they joined the line, renege and return &#8220 ; ; tomorrow&#8221 ; ; and mail the gift. A participant began the experiment with a fixed endowment which diminished with time spent in line and also if s/he balked or reneged. We found that participants were more likely to join and stay until served if they either knew that clock time (or some function of clock time) would be provided, or if they were explicitly told the expected waiting time before they would be served, or both. The impact of explicit information was much greater than that of clock time, and clock time in turn was more effective than any function of clock time such as elapsed time. Thus, to prevent balking and reneging managers should consider providing customers accurate time estimates of the expected wait A participant&#8217 ; ; s perception of the time before s/he would be served was a key covariate in analyzing the experimental data. We found that participants were of two types: approximately 60% underestimated waiting time in long lines but overestimated it in short lines, and 40% did the opposite. Our measurement of anticipated waiting time is in contrast to prior studies of subjective time in which participants are asked to estimate how long they have waited after service is completed, and have consistently overestimated this time. Queuing theory shows that participants should never renege, and they should use a threshold rule for balking. For the parameters in our experiment, the threshold was 7 &#8211 ; ; participants should balk for line lengths greater than or equal to 7. This threshold value is based on the expected line length they will face when they return to the queue for service on the &#8220 ; ; next day&#8221 ; ; plus the switching cost. Evaluated against this benchmark, decision-making on the whole is quite good. Reneging is rare &#8211 ; ; it occurs on only 4.8% of experimental occasions. Most participants use a consistent threshold rule of balking with a modal value of 8, compared to the benchmark of 7. However, their estimate of the line length on the next day is extremely accurate, suggesting that participants are systematically overestimating the switching cost. Thus managers who wish to switch customers for example from a peak load to a slack load period may have to provide significant incentives.

decision making; services marketing

nije evidentirano