UTILIZAÇÃO DA DISTRIBUIÇÃO DE ERLANG PARA RENTABILIZAR OS SERVIÇOS DE TELEFONIA. (ESTUDO DE CASO NA CENTRAL TELEFÓNICA DA ZAP FIBRA – LUANDA)

Abstract

In this article, we will apply the Erlang distribution to calculate the customer arrival and service distribution regarding the queue system, which is related to the efficiency and capacity of the telephone center under analysis of Zap Fibra – Launda. The objective of this article is to use the Erlang distribuition to optimize or increase the efficiency of operators at a telephone center with 6 workstations, aiming to increase customer satisfaction by reducing their wainting time in line for service, thus increasing the capacity of the telephone center and its profits. The waiting time process is not entirely predictable; for this reason, a way to represent this variability needs to be found, and in this case, the Poisson and exponential distributions will be used. To vary out the work, a quantitative and qualitative study was distributions will be used. To carry out the work, a quantitative and qualitative study was chosen, with descriptive analytical purposes. In order for this to be possible, it was necessary to collect data related to the number of customers who called the Zap Fibra – Luanda customer support number over the course of 30 days, with the aim of requesting information and activation of Zap Fibra services. The data used for the study were collected through interviews with employees of Zap Fibra´s call center, where a total of six (6) employees were selected. Four (4) males and two (2) females, with an age range between 22 and 40 years old. During our research, we chose to use the interview technique as it is explicitly a qualitative technique that allows us to collect and analyze the interviewees opinions on various aspects of life at a personal, family, social, and economic level. And for the treatment of the collected information, the electronic support Nvivo 12 was used, one of the most widely used software for conducting qualitative studies. In summary, the application of the Erlang distribution allowed us to conclude that, with an average of 5 calls per hour, assuming they arrive at the same moment, only 4 operators can be maintained. Because, with this number of available operators, the probability of a queue will be only 40% and the level of services provided will be 52%, which is significantly positive for the intentions of the analyzed company.