Revenue Forecasting Models for Hotel Management

Abstract

The Holt-Winters model works well with data that contain seasonality and trend ... to optimize revenue, companies need a formal forecasting process ... hotels geared toward skiing and golf are very much affected by a change in the weather. Every organization operates under some kind of uncertainty. Forecasting helps to manage uncertainty. There are two main approaches to forecasting. One is based on the cause-andeffect approach where a forecast is based on the value of factors that drive the event. The other is based on the extrapolation of historical data. Between the two, the former approach is a little more difficult than the latter. However, regardless of forecasting method used, all organizations benefit from forecasting. SKI RESORT AND CONFERENCE CENTER The hotel under study is a ski resort and conference center located in southwestern New York state. The hotel offers several options for lodging with 108 rooms, including condominiums that are designated for use as rentals, ownership, and timeshares. The hotel alters its market focus throughout the year because of the seasonal nature of its business. In the winter months, the market represents primarily the ski business, because this hotel is the only ski resort within a 60mile radius. In the spring, summer, and fall months, government agencies and corporations hold business conferences in the area. Since there is a wide selection of conference facilities in the area, the hotel normally lowers its prices during that time to remain competitive. The management feels that the hotel is not optimizing its revenue, and the way to remedy this is to implement a good forecasting system. The management believes that revenue forecasts are the major input to revenue management. But with no precise forecasts, a revenue management system does not give optimum rate and room availability. Indeed, a recent article showed that a 10% increase in forecast accuracy in the airline industry improved revenue anywhere from 0.5% to 3.0%. Revenue projections help determine cash flow and help keep the hotel at an operational level. Revenue forecasting can be done on a weekly basis, enabling the management to budget most efficiently the room rates as well length of stays. We apply this model to the hotel mentioned earlier to maximize its revenue. Here, we use 54 weeks of data for forecasting revenue. WHY FORECASTS? Like any other industry, forecasting is essential in the hotel industry. Without forecasting, hotels would base pricing and length of stay on gut feelings instead of on actual demand for the market. Although forecasts are seldom 100% accurate, they offer decision-makers valuable inputs in making decisions. Hotels mostly use the occupancyforecasting model. In order to advise functional departments how busy they will be, a variety of specifics has to be analyzed. Arrival and departure patterns are needed for budgeting appropriate resources, employees, and work shifts to make the most of their resources. Seasonality is a major part of forecasting, particularly in the hotel business. The times of the year when people travel the most are spring break, summer, and Labor Day weekend. These are the busiest times for hotels. Having additional staff and supplies for those times is essential for ensuring that guests are fully satisfied and their stay is optimal. Another thing to forecast is room type preference. If more guests prefer double occupancy instead of single occupancy, it would be most beneficial to have more of that type of room available. Demand and revenue forecasting are most important for managing business. Demand forecasting establishes selling procedures and criteria for the revenue forecast. This information should be gathered on a daily basis to track the stay patterns for future forecasting efforts. In this article, we use revenue forecasting since it provides information to managers about occupancy levels, nightly average rates, and sales estimates. …