Student Study Notes - Chapter 10

 

 

The Importance of Accurate Forecasts

• One of the first questions restaurateurs and hoteliers must ask themselves is very simple: “How many guests will we serve today? This week? This year?” The answers to questions such as these are critical, since these guests will provide the revenue from which basic operating expenses will be paid.
• Labor required to serve the guests is also determined based on the manager’s “best guess” of the projected number of customers to be served and what these guests will buy.
• Forecasts of future revenues are normally based on a careful recording of previous sales, since what has happened in the past in an operation is usually the best predictor of what will happen in that same operation in the future.
• Finally, operating, cash, and capital budgets (see Chapter 11) cannot be prepared unless an operator knows the amount of revenue upon which these budgets should be based.
• In its simplest case, sales are the dollar amount of revenue collected during some predetermined time period. When used in this manner, sales and revenue are interchangeable terms. A distinction is made in the hospitality industry between sales (revenue) and sales volume, which is the number of units sold.

• With accurate sales records, a sales history can be developed for each foodservice outlet you operate, and better decisions will be reached with regard to planning for each unit’s operation. In a similar manner, knowing the number of hotel rooms to be sold allows hoteliers to staff appropriate numbers of room attendants, laundry workers, front desk staff, and food and beverage employees. Because operating costs are typically well-known, the identification of sales levels naturally leads to improved identification of estimated expenses.
• Managers utilizing forecasts and forecast data understand some basic truths about forecasts. These include:

• Forecasts involve the future.  Managers can make long-term forecasts and simply modify them as more precise short-term forecast data become available.

2.   Forecasts rely on historical data.  In some cases, historical information may be missing, or the accuracy of historical data that does exist may sometimes be questionable. In other cases, unusual events may serve to make historical data misleading. For these reasons, managers utilize historical data, but then carefully add their own estimates and predictions about how that data will vary or be repeated in the future.

3.   Forecasts are best utilized as a “guide.”  Forecasts should be accurate, be based upon the best information available, and be free from unwarranted influence or bias. When forecasts have these characteristics, they can indeed serve as useful managerial guides.

Forecast Methodology

• If only historical data was used to predict future data, forecasting (at least for operations that are already open) would seem to be simple. In fact, in most cases, variations from revenue forecasts are likely to occur.
• When a variation does occur, experienced managers know that some of it can be predicted. Assume that a restaurant has been, for the past several months, experiencing a 10% increase in sales this year when compared to the same period last year. This trend, or directional movement of data over time, of increased sales may be very likely to continue
• Several types of trends may occur that can help a hospitality manager forecast revenues.  A seasonal trend, or a data pattern change due to seasonal fluctuations, can be fairly accurately predicted because it will happen every year. Cyclical trends tend to be longer than a period of one year and might occur due to a product’s life cycle, such as the downturn of revenues after the “new” wears off of a trendy concept.
• Finally there can simply be random variation. This variation appears to occur on a totally unpredictable basis. Upon closer examination however, some random events can be identified. The ultimate goal you should set for yourself as a professional hospitality manager responsible for forecasting sales revenues, expenses, or both is to better understand, and thus actually be able to predict, as much of this random variation as possible.

 

Forecasting Restaurant Revenues

• For operating restaurants, accurate sales histories are essential for forecasting future sales. A sales history is the systematic recording of all sales achieved during a predetermined time period, and is the foundation of an accurate sales forecast. When you predict the number of guests you will serve and the revenues they will generate in a given future time period, you have created a sales forecast.
• The simplest type of sales history records revenue only. The sales history format shown in Figure 10.1 is typical for recording sales revenue on a weekly basis.
• Some hospitality managers do not have the ability to consider sales in terms of revenue generated (see Figure 10.2 for an example of the type of sales history used when no cash sales are typically reported). This non-revenue sales history approach is often used where knowledge of the number of actual guests served during a given period is critical for planning purposes.
• For most restaurant operations, sales revenue consists of the number of guests served as well as how much each of those guests spend. Sales histories can be developed to track the number of guests served as well as to compute average sales per guest, a term commonly known as check average (see Chapter 6). Recall that the formula for average sales per guest is:

 

• Maintaining histories of average sales per guest is valuable if they are recorded as weighted averages. The reason why can be seen easily by analyzing the data in Figure 10.3.
• The use of a simple average, or the value arrived at by adding the quantities in a series and dividing that sum by the number of items in the series, to calculate average sales per guest would not be correct.
• In fact, the weighted average sales per guest, or the value arrived at by dividing the total amount guests spend by the total number of guests served, should be used.
• In most cases, sales histories should be kept for a period of at least two years. This allows managers to have a good sense of what has happened in their business in the recent past as well as the most recent time periods.

 

Forecasting Future Revenues

• To learn how managers use sales histories to forecast future sales, consider the sales history shown in figure 10.4.

 

Go Figure!                                                                                                                          

The dollar variance for sales for the period is computed as follows:

The percentage variance for the period is computed as follows:

Another way to compute the percentage variance is to use a math shortcut, as follows:

                                                                                                                                             

It is important to understand that sales histories do not record the reasons for increases or decreases in revenue. Knowing those is the job of the operation’s manager.  See Figure 10.5 for an illustration of how a planning sheet for the next period could be developed.

Go Figure!                                                                                                                          

For the next period total, the sales forecast is calculated as follows:

An alternative way to compute the sales forecast is to use a math shortcut, as follows:

                                                                                                                                             

Forecasting Future Guest Counts

• Using the same techniques employed in estimating increases in sales, the noncash operator or any manager interested in guest counts can estimate increases or decreases in the number of guests served (see Figure 10.6).
• See Figure 10.7 for an illustration of how a planning sheet for predicting the guest count for the next period could be developed.

 

Go Figure!                                                                                                                          

For the next period total, guest count forecast is calculated as follows:

This process can be simplified by using a math shortcut, as follows:

                                                                                                                                             

Forecasting Future Average Sales per Guest

• Average sales per guest (check average) is simply the amount of money an average guest spends during a visit. The same formula is used to forecast average sales per guest as was used in forecasting total revenue and guest counts. Therefore, using data taken from an operation’s sales history, the following formula is employed:

 

• Alternatively, you can compute average sales per guest using the data collected from sales forecasts divided by the data collected from guest count forecasts. Figure 10.8 illustrates how the data from Figures 10.5 and 10.7 can be combined to compute a sales per guest forecast.

 
Go Figure!                                                                                                                          

For the second period total, the average sales per guest forecast is determined as follows:

                                                                                                                                             

• Sales histories, regardless of how well they have been developed and maintained, are not, when used alone, sufficient to accurately predict future sales. The operation’s managers must have knowledge of potential price changes, new competitors, facility renovations, and improved selling programs to name just a few of the many factors that must be considered when predicting future revenues.

 

Forecasting Hotel Revenues

• Accurate and useful forecasts are even more important to hotels than they are to restaurants. This is so because, unlike restaurateurs, hoteliers are most often held accountable not only for controlling costs, but also for the short-term management of revenues via yield management and other RevPAR maximization techniques. In addition, hotel room rates (unlike most restaurant menu prices) are likely to be adjusted daily, weekly, and monthly based upon a hotel’s forecast of future demand for its hotel rooms.
• Forecasts that are 100% accurate are rare. Forecasts that are consistently or significantly in error, however, will ultimately result in significant financial or operational difficulty for a hotel. This is true whether the forecasts are too high or too low. Forecasts that are implausibly too high:

• Cause unrealistic expectations by the hotel’s owners
• Increase feelings of frustration by affected staff when forecasted volume levels are not attained
• Produce budgeting/ spending errors by overstating anticipated revenues
• Result in impractical and overly aggressive room rate determinations. When forecasts are excessively high, room rates may be set too high (see Chapter 8).

• Alternatively, forecasts which are consistently and unrealistically too low:

• Lead management to believe it is actually performing at higher levels of room sales
• Undermine the credibility of the forecaster(s) because of the suspicion that actual forecast variation is due to low-balling (intentionally underestimating) the forecasts
• Result in impractical and under-aggressive room rate determinations (rates too low)

• With experience, those responsible for forecasting demand for a hotel’s rooms can produce room forecasts that will be within 1% to 5% of actual hotel room revenues.
• Accurate demand forecasts are not an end, but rather, as you learned in Chapter 8, a means of improving the effectiveness of establishing room rates. Demand forecasts predict periods in which demand will be generally high or low based on expected fluctuations in occupancy. Types of hotel forecasts can be found in Figure 10.9.
• While restaurateurs rely primarily on historical (and in some cases predictive future data), to estimate revenues and guests to be served, hoteliers must rely on a combination of historical, current, and future data to accurately forecast and manage room demand.
• Historical data, of course, refers to events that have happened in the past. Current data is related to events that are entered into a hotel’s property management system (PMS) but have yet to occur. Accurate current data reflects confirmed (but as of yet not historic) information about the demand for hotel rooms. Future data is that information which is related to events that have yet to occur and are not currently be found in the hotel’s PMS database.
• Hoteliers create occupancy forecasts in a manner similar to how they compute actual occupancy percentage. That is:

 

• Increasingly, PMSs are designed to include forecasting programs or components. While PMS forecasting modules may indeed “guide” them, local property managers themselves will ultimately make the best occupancy forecasts for their own hotels because they:

• Understand the unique property features affecting demand for their own hotels
• Know about special city-wide events in the area that affect demand
• Have an understanding of the demand for competitive hotel properties in the area
• Can factor in the opening or closing of competitive hotels in the area
• Can include factors such as weather, road construction, and seasonality in their demand assessments
• Can adjust forecasts very quickly when faced with significant demand-affecting events (i.e., power outages and airport or highway closings)

 

Historical Data

• One of the best ways for existing hotels to predict future room demand is by examining historical demand. Managers combine their own skills and experience with relevant historical (and other) data when creating usable demand forecasts. In order to create these forecasts, you must first understand the following terms:

• Stayover:  A guest that is not scheduled to check out of the hotel on the day his or her room status is assessed. That is, the guest will be staying and using the room(s) for at least one more day.
• No-show: A guest who makes a room reservation but fails to cancel the reservation (or arrive at the hotel) on the date of the reservation.
• Early Departure: A guest who checks out of the hotel before his or her originally scheduled check-out date.
• Overstay: A guest who checks out of the hotel after his or her originally scheduled check-out date.

• Figure 10.10 shows the method used to compute a single day’s occupancy forecast for a 300 room hotel. In an actual hotel setting, room usage and availability would be forecast by individual room type, as well as for the total number of hotel rooms available. The procedures and steps are the same when forecasting room type availability and/or total room availability.
• The number of hotel rooms available, the number of out of order rooms, the number of stayovers, and the number of reservations currently booked are all data that resides in the PMS. Data for the three forecast adjustments (no-shows, early departures, and overstays), however, describe events that will occur in the future, and thus “real” data on them does not exist. These numbers must be forecast by managers after carefully tracking the hotel’s historical data related to them.

 

Using Current and Future Data

• Historical data in the PMS is very valuable because room demand often follows fairly predictable patterns. However, the use of historical data alone is, most often, a very poor way in which to forecast room demand. Current and future data must also be assessed.
• In the situation presented in Figure 10.11, estimated demand based on the historical data would be far short of the demand indicated by the current data. On-the-books is the term hoteliers use to describe current data and it is used in reference to guest reservations. The term originated in the days when hotel reservation data was stored in a bound reservations book.
• Future data is the final type of information needed to assist hoteliers in accurately forecasting demand. In fact, most hoteliers agree that a manager’s ability to accurately assess this information is the most critical determiner of an accurate demand forecast.
• As can be seen from the examples given in the text, if hoteliers are to make accurate forecasts and properly price their rooms, historical, current, and future data must all be carefully considered. This is so because occupancy forecasting is not simply a matter of identifying the number of hotel rooms that may be sold, but rather it is a multifaceted process that consists of four essential activities that include:

• Generating the demand forecast
• Establishing a room rate strategy
• Monitoring reservation activity reports
• Modifying room rate pricing strategies (if warranted)

• For hoteliers (unlike restaurateurs), pricing decisions naturally follow forecast development. Thus, accurate demand forecasts will profoundly affect a professional hotelier’s room pricing decisions.
• Only by creating an accurate forecast can a hotel know when room demand is strong or weak enough to dictate significant changes in pricing strategies and thus affect the procedures and tactics designed to help a hotel achieve its RevPAR and revenue per occupied room (RevPOR) goals.

 

Utilizing Trend Lines in Forecasting

• Sophisticated business forecasters in many fields have found that the application of advanced mathematical models can often result in forecasts which are much more accurate than those derived from approaches which do not utilize such advanced formulas.
• Fortunately, such formulas exist in somewhat basic forms and are easy for most managers to use. One such popular formula produces a trend line, which is a graphical representation of trends in data that you can use to make predictions about the future. For the purpose of this chapter, a trend line will be used to forecast future sales.
• This analysis is also called a regression analysis. A regression analysis estimates an activity (dependent variable - forecasted sales in this case) based on other known activities (independent variables - past sales in this case). By using a regression analysis, you can extend a trend line in a chart beyond the actual known data for the purpose of predicting future (unknown) data values.
• Once the annual sales data for several fiscal years has been collected (see Figure 10.12) a line graph can be created using Excel (see Figure 10.13).
• A trend line (future prediction) can be created, using the baseline data (known data). All managers creating trend lines must ensure that:
• There is enough data to show a meaningful trend. Insufficient baseline data will likely skew results.
• The data is entered into the spreadsheet from earliest (oldest) to most recent (newest).
• No data is missing. If data is unavailable for a period, an estimate must be entered.
• All periods are for comparable amounts of time, such as weeks, months, or, as is the case in this example, fiscal years.

• If all of the above items are satisfactory, a trend line can be created to predict future sales levels using Excel (see Figure 10.14).