Automating Forecasting to Reduce Overall Food Costs
In a blog post we wrote a while ago, we highlighted the 7 key measures of performance for foodservice leaders. At the top of the list was reducing overall food waste. As cost management continues to drive the healthcare environment, reducing overall food waste should be a top priority for all foodservice operators. LeanPath estimates that an astonishing 4-10% of food purchased in US foodservice operations is thrown out before even reaching a consumer. Much of this waste is a direct result of overproduction.
The first key to reducing over production is using an accurate, reliable forecasting model to identify production needs. Manual forecasting methods can miss historical sales trends by neglecting to consider actual sales counts and menu mix in predicting future menu needs. Worse yet, manual forecasted counts are often padded by two or more people resulting in greatly overestimating actual needs.
Accurate forecasting must consider menu mix when calculating production needs. Menu mix can be defined as:
- The relative fraction or percentage of sales that each menu item contributes to the whole
- The ranking of each menu item by customer preference (popularity) by meal period
- The total number of an individual item sold/selected divided by the total number of all items sold/selected
The menu mix can change dramatically based on which menu items are being served at the same time. Consider for example, if Lasagna and Liver are being served together, 90% of customers may choose the Lasagna. However, if Lasagna and Chicken are served together, perhaps only 50% of customers will choose the Lasagna. If this shift in menu mix is not considered when forecasting, the Lasagna could be greatly over produced.
A computerized model can assist operators in more accurately predicting forecasts by reviewing post sales history for the menu mix of items being served. This is especially advantageous if the software application integrates with point-of-sale figures so that the data does not have to be manually updated.
Of course, in addition to menu mix, there are many other external factors that can impact forecast volumes, especially in retail establishments. Some examples include weather, pay day cycles, or special events that dramatically increase or decrease customer counts. Therefore, it is imperative that a computerized forecasting model also provides mechanisms to accommodate for these factors in order to fine-tune the forecast figures.
Bottom line, getting a handle on forecast predictions is the first step to reducing food costs associated with over production. From there, follow through to ensure that food is ordered and produced in accordance with the menu needs.
Article by: Heather Johnson, Hospitality Suite Product Manager; Fusion, 4th Quarter, 2013