L. L. Bean Improves Call-Center Forecasting

Two forecasting models for use at L. L. Bean, Inc., a widely known retailer of high-quality outdoor goods and apparel, are developed and implemented and the improved precision is estimated to save $300,000 annually through enhanced scheduling efficiency.

Abstract

We developed and implemented two forecasting models for use at L. L. Bean, Inc., a widely known retailer of high-quality outdoor goods and apparel. The models forecast calls incoming to L. L. Bean’s call center so that efficient staffing schedules for telephone agents can be produced two weeks in advance. We used the ARIMA/transfer function methodology to model these time series data since they exhibit seasonal patterns but are strongly influenced by independent variables, including holiday and advertising interventions. The improved precision of our models is estimated to save $300,000 annually through enhanced scheduling efficiency.