SIMULATION AND EXPERIMENTAL DESIGN APPLIED TO SIZING SUPERMARKET CASHIERS IN COLOMBIA Jorge A. Alvarado Luis M. Pulido Carrera 7 40-62 Edificio José Gabriel Maldonado Piso 3 Pontificia Universidad Javeriana Bogotá, COLOMBIA ABSTRACT A framework for Colombian supermarkets challenges in order to find an adequate number of cashiers and baggers was developed, translated to a simulation model using Promodel® and proved through an experiment varying reg- ister item time, number of cashiers and number of baggers for eight real Colombian supermarkets. The framework proved to be successful in finding a cashier-bagger combi- nation for average waiting and system times, but not pow- erful enough for service promises involving all of the most part of the clients due to variability of maximum times. An interaction among cashiers and baggers combination was found that makes valuable to increase the number of bag- gers only for some specific number of cashiers working. 1 INTRODUCTION Finding an adequate number of cashiers for a supermarket could be a hard-to-solve problem for a retailer involved in this business. As Dick Larson stated, (Parker 2003) , “In supermarkets, the real state is too valuable. If anyone can solve the supermarket queuing problem- the way banks and airlines have done- that person will become rich and fa- mous”. Although the great efforts supermarket chains proba- bly have done, there are few public literature available ad- dressing that specific problem. Melachrinodius y Olafsson (Melachrinodius and Olafsson 1992) proposed a cashier scheduling taking into consideration only hourly labor de- mand constraints and availability of labor constraints for each day of the week, in order to minimize the number of cashiers. Recently, (Melachrinodius and Min 2008) that model was improved adding into consideration order sizes and customers check-out station selection method, fixing an acceptable level of customer service to be met, and re- fining the objective function. Other attempts to model and simulate services have addressed different problems, such as banks and call centers (Saltzman and Mehrotra 2001); (Deutsch and Mabert 1980), but didn´t attack the specific supermarket problem with their particular difficulties. The research objective was to create a general frame- work for Colombian supermarkets in order to face the de- cision about the minimum number of cashiers and baggers needed given a service level, put the model in a simulation environment and prove the effect of the main factors de- fined and their second level interactions through experi- mental design based in real data. The rest of the paper is ordered as follows: The second section shows the challenges and specific factors to be taken into account in the supermarket environment and in the Colombian context; the third part explains the data col- lection process and explains the experimental design per- formed; the fourth heading discuss the results obtained; and the fifth part has the conclusions, remarks and prospec- tive of the research work. 2 THE MODEL After reviewing the literature available and interviewing three cashiers supervisors in different supermarket chains, three factors were chosen to be simulated: day-of-the- month scenario, supermarket size and item register time. Two supermarket supervisor decisions were evaluated: number of cashiers and number of baggers. In addition, a new cashier selection method (by the customer) was devel- oped. All the supermarkets have regular cashiers and a fast cash for customers who has less than eleven items to pay. The service promise was to serve 98% of the customers in less than 7 minutes. 2.1 Cashier selection method A customer doesn´t select a line in a supermarket at ran- dom. For instance, researchers have found that the appear- ance of the line does matter in customer perception (Hornik 1984) Usually, a customer evaluates the queue length in his decision. But in a supermarket, not only the 1356 978-1-4244-2708-6/08/$25.00 ©2008 IEEE Proceedings of the 2008 Winter Simulation Conference S. J. Mason, R. R. Hill, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds.