Group Forecasting Accuracy in Hotels
研究了北美大型连锁酒店约90家酒店的团体预订预测数据，发现预测存在正向偏差，提前两个月平均绝对百分比误差为40%，提前一个月为30%，入住当天降至10-15%。大型酒店、依赖团体业务的酒店及频繁更新预测的酒店准确性更高。
Yield management helps hotels more profitably manage the capacity of their rooms.Hotels tend to have two types of business: transient and group.Yield management research and systems have been designed for transient business in which the group forecast is taken as a given.In this research, forecast data from approximately 90 hotels of a large North American hotel chain were used to determine the accuracy of group forecasts and to identify factors associated with accurate forecasts.Forecasts showed a positive bias and had a mean absolute percentage error (MAPE) of 40% at two months before arrival; 30% at one month before arrival; and 10-15% on the day of arrival.Larger hotels, hotels with a higher dependence on group business, and hotels that updated their forecasts frequently during the month before arrival had more accurate forecasts.