Research On Traffic Forecast Optimization Of K Customer Service Department Of State Grid Customer Service Center

With the transformation and development of social economy,more and more industries have established their own call centers.While providing customers with basic services such as business consultation,business handling etc.,they also use the huge customer information holdings of call centers to carry out market marketing promotion,which is an effective strategy for enterprises with high efficiency and low cost marketing.The key performance indicator of a call center is the reception rate,which can directly represent the basic service ability of a call center,and can also reflect the management and operation of a call center.And the accuracy of traffic prediction is the most important factor to affect the response rate.It is reported that at present,many call centers do not perform well in the accuracy of traffic prediction,not only lack of effective traffic prediction model for quantitative prediction,but also lack of rich operation management experience to predict traffic.Therefore,the research on the optimization of traffic prediction has far-reaching significance for the development of the whole call center industry.Traffic prediction of ascension to call center of artificial cost reduction and improve operational management level in the call center has a great role in promoting.First of all,the annual traffic forecast has important guiding significance for the recruitment and new staff training arrangement of human resources department.Secondly,the forecast of the daily traffic volume of 48 points in half an hour per month plays a decisive supporting role for the monthly schedule and daily training arrangement.The purpose of this paper is to improve the traffic prediction accuracy of K Customer Service Department.Firstly,it analyzes the traffic characteristics of 95598 and the shortcomings in the process and model of the current traffic prediction in K customer service department.Then,it determines two optimization levels of traffic prediction: First,process optimization is carried out from five aspects of Define Measure Analyze Improve Control through DMAIC model.The second is to design the traffic prediction model based on the cubic exponential smoothing algorithm,and make quantitative prediction from the aspects of data cleaning influence traffic factor coefficient confirmation,traffic data restoration,model smoothing coefficient determination and data iteration prediction result correction,so as to improve the traffic prediction accuracy of K customer service department and reduce the operation cost;then,the paper compares and analyzes the improvement of various indexes such as the accuracy of traffic forecast in K customer service Department before and after optimization,such as the human service call rate scheduling fitting time utilization rate and emergency response times,and tests the effectiveness of the process optimization and traffic prediction model.Future traffic prediction work step by step can be online,reduce artificial factors of traffic forecasting precision,at the same time through programming to build corresponding traffic prediction system,will be more accurate prediction model used in the operation management,make traffic prediction work more lean and precision at the same time,the modular universal part of traffic forecast model and customization according to different regions of the parameter selection,increase traffic prediction model can extension.