ABSTRACT
Forecasting is a cornerstone of hotel revenue management. Its quality has direct impact on hotels’ ability to generate higher revenues and profits. A crucial yet often overlooked aspect is the intricate nature of the forecasting accuracy measures used to evaluate forecasting performance. Several forecast accuracy measures have been proposed in the literature. In the next section, for every measure a formula is given together with a step-by-step computation example. The geometric mean relative absolute error compares the geometric average of the forecasting error with the geometric average of the benchmark error. In order to compare Group and Transient segments, scale-independent forecasting accuracy measures have to be used. Forecasting accuracy measure can be used for two distinct purposes: in the phase of selecting a forecasting model to screen out a forecasting method that is not appropriate; once the forecasting model is selected, forecasting accuracy measures to monitor the forecasting model’s performance.
