Contact center productivity improvement based on predictive analytics

EasyChair Preprint 2097

Abstract

The overall performance of a contact center is often measured through its ASA (Average Speed of Answer), abandonment rate and Service Level (SL). This study focuses on the predicting ASA as a function of call volumes, AHT, occupancy of agents, number of productive Full Time Equivalents (FTE) and Off Phone Activities % (OPA) and analyzing the impact of each of these parameters on ASA through sensitivity analysis. An ensembled model for ASA prediction was created using multi-variate time series prediction algorithms like ARIMAX and neural network predictor based on 3 years of data from a contact center in US

Keyphrases: Average Speed of Answer, Contact Center, capacity planning, time series analysis

@booklet{EasyChair:2097,
  author    = {Rishabh Tyagi and Sandeep Bhattacharya},
  title     = {Contact center productivity improvement based on predictive analytics},
  howpublished = {EasyChair Preprint 2097},
  year      = {EasyChair, 2019}}