What does hotel forecasting specifically consist of?
Let's first look at a definition of Forecasting. As we explained in our article on Revenue Management, an important phase of the revenue management process is the Hotel Forecasting (or Forecast). Hotel forecasting is a forecast of demand and revenues for future dates based on the analysis of historical data, bookings on the book, the market and its competitors. In this way, the revenue manager will be able to optimize hotel revenues by defining an offer that meets the needs of the demand.
Forecasting for a given future date can be of three types:
- Occupancy Forecasting: the revenue manager makes a forecast of the occupancy level;
- Demand Forecasting: the revenue manager obtains the total or “unconstrained” demand (“unconstrained demand”, ie the level of occupation that does not consider the accommodation capacity of the structure);
- Revenue Forecasting: the revenue manager makes an estimate of the revenues that will be generated.
You can apply various forecast models for hotel forecasting that fall into three main categories:
Historical Booking Models
These hotel forecast models do not take into account existing bookings, but only analyze historical data relating to a specific date with the aim of identifying a cyclical trend or a seasonality: in this way it will be possible to predict the final occupancy relating to a specific future date. Some models that fall into this category are Same Day Last Year, Moving Averages, Exponential Smoothing, and other time series.
Advanced Booking Models
Advanced Booking Models analyze existing bookings and their evolution over time with respect to a specific date. This category includes the synthetic booking curve and the pick-up, a model that analyzes the pace of bookings.
Combined Models
They are a combination of the other two categories: some examples are the Regression Weighted Averages of the predictions obtained with the Historical and Advanced models.
The revenue manager will analyze numerous variables such as historical data, occupancy trends, the presence of events for specific dates, information relating to bookings and fares, prices charged by competitors or flight bookings for his destination.
All information relating to the market and competitors can be obtained through a hotel benchmarking platform such as HBenchmark, which returns historical and forecast price and occupancy data relating to the destination and reports all events that may have a significant impact on the accommodation facilities of the territory. By applying the forecasting models described above to all these data, the revenue manager will obtain valuable suggestions regarding the allocation of rooms and the pricing strategy to be applied.
