What is a revenue driver? 

A revenue driver is what drives the revenue of a department. For example, if more guests are staying at a hotel, the restaurant can expect to generate more revenue. There is often more than one driver that impacts revenue in a department. 

A driver-based forecast uses these drivers to forecast for the department. For example, a restaurant may use Guest nights as a driver to forecast the number of covers. 

How to review and update the driver 

If set-up has selected to use PMI machine learning algorithms, the driver detection will be automated based on seasons and day of the week. It will create different drivers per segment and, if units/covers are used, separate drivers for these. This is recommended for properties with at least 1 year of historic data in PMI. It is possible to use ML (Machine Learning) drivers with only 5 weeks of history, but the accuracy will increase as the amount of historic data available increases. 

To view the drivers being used: 

1. Go to tools. Select ’Driver based’  

2. Use the graph to review the results and evaluate the usefulness of the drivers. You cannot make any edits to drivers when ML drivers are activated. Consider if your department could achieve a more accurate forecast with a manual driver.
   1. ML drivers are detecting correlations between the department’s historical covers/units/revenue and other departments/segments (sources). There can be more than one source per day of the week which can make it difficult for humans to understand the correlations. We therefore recommend using the graph for a general evaluation of the accuracy of the drivers, instead of checking the sources in the table. 

Manual drivers are usually used for new properties with less than 1 year of historical data available, or for properties that have recently been through significant changes that make historic patterns an unreliable source for future.  

To create a manual driver:  

1. Go to Tools and select Driver based
2. Use the drop-down menus to define which segment (if enabled) you want to build a driver for.  
3. Choose the source. You can also define which segments within the source you want to build the driver from.
   1. Choose if you want the calculations by weekdays or just as a total (unticked radio box). 
4. After you have defined settings, click on the + sign and the rule for this segment is moved up in the rate grid list. You may build several drivers for one segment or department. 
5. After all the rules have been created, choose if you want the entire forecast to be based on the rule. 100% means everything, but if you want to revise e.g., 30% of the revenue manually, change the cell with percentage to 70. If you have a positive change in the revenue with 10% on specific days or in total, you can change the percentage to 110%. 
6. Applicable for covers only: Define capture ratio for both cover/units and rates. You may untick both these buttons if desired. 
7. Applicable for covers only: An easy way of letting the system help you is by clicking on one of the two blue buttons, From Budget or From History.  From Budget means that PMI splits the budgeted data per day based on last year’s actual figures. It will suggest the same average check, but different number of covers each day. In other words, PMI calculates the budgeted capture ratio of occupancy; e.g., if your budget is 100 covers and guest nights are budgeted at 1000, the capture ratio would be 10%. This is usually recommended to new hotels that don’t have any historical data. 
8. Applicable for covers only: From History means that PMI calculates each departmental segment individually, giving you different averages for each segment, and different average rates, based on last year’s actual figures. For example, if you had 10 guests on average, last year on Mondays, and the guest night occupancy was, on average, 100, PMI will calculate 10% of guest nights as the capture ratio for this period. 
9. You may calculate for just some of the rules as well. Use the radio button next to the rule to activate or deactivate for calculation. 
10. You may manually revise the ratios directly in the cells at any time. You can also make a New Period for different seasons that influence the average revenue. For example, summer is a slow season with guests spending less money, on average, in the restaurant, than in the autumn with conference guests on average spending more. 
11. Make sure that you save the changes.