|
|
Case Study – Call Center Operations
Call Centers have become the principal interface between businesses and their customers, particularly in the service and finance industry sectors. Customer experience of a business can be irreparably damaged by inefficient or impersonal call center responses. It is said that after only two frustrated calls, 85% of those customers move their business away. (A 5% improvement in customer retention improves profitability by 25 to 100% - Bain & Co.).
Call center efficiency is affected by many factors, not least of which are fluctuations in in-coming call volumes. Traditionally, as call volumes have increased, more staff has been hired. The limitations being the availability of suitable people and the associated training and other direct costs. Our client recognized these difficulties and sought alternative solutions.
Modern Analytics applied sophisticated time series analytical models to predict the likely volume of in-coming calls that would be received each hour, each day for six months in advance by the client's 500-person Call Center. Now, by knowing well in advance the likely level of inbound and outbound call volumes per hour, the client can adjust its staffing and calls distribution to meet and exceed its customer response goals without additional hiring.
These highly detailed Call Center predictions, for up to six months in advance, utilized Auto-Regressive Integrated Moving Average (ARIMA), Spectral Analysis and Recurrent Neural Networks time series models to predict daily call volumes and call distributions. The results were impressive; 95% of the forecasts were within 3% of the actual values. The client experienced significant labor cost savings, an increase in overall response efficiency and greatly improved customer satisfaction.
|
|
