4 Major Differences in Restaurant Forecasting Systems
Using the right restaurant forecasting system can save an enormous amount of time and money, and prevent the labor and food shortages that disappoint customers. Systems run the gamut from spreadsheets created in-house, to modules fully integrated into enterprise solutions.
Following are four areas in which forecasting systems can differ substantially, and which can deeply influence their effectiveness.
1. A Choice of Metrics
Forecasts are typically generated using historical sales or guest-count data as a starting point, and then adjusted as necessary for current variables. For example, a system might go through the following process in creating a forecast:
- Review the sales trends for the previous 4 weeks
- Determine whether business is growing or falling, and by how much
- Use last week’s sales as a base model, then adjust to reflect the trend
- Use the result as the coming week’s forecast, distributed across the various days, day-parts and hours according to the average distribution for the past 4 weeks.
An easy to use forecasting tool gives the restaurant a planning expert that’s always on the job.
While using the previous 4 weeks’ sales data might be right for some operators, others may want to use the previous year’s sales for the same period, to help account for annual variables like holidays, events, or school vacations. Instead of using sales, some may want to use guest counts, or even transaction counts to forecast, which help to minimize the revenue fluctuations that price changes create. And still others may want to compare year-over-year trends (scaling last year’s sales up or down according to recent trends). The point is that the forecasting system should give operators the flexibility to easily choose the metrics that work best for their operation.
2. Integration with other systems
You may have seen it… one manager creates a forecast in a spreadsheet, and another places vendor orders without first seeing that forecast. Trouble ensues. For maximum efficiency, all actions regarding food and labor use in the restaurant should be based on a single forecast for the coming period. The simplest and most consistent way to do that is to have the forecasting system integrated with the ordering and inventory, labor scheduling, and prep/production systems. In that case, once a forecast is entered:
- A fully integrated ordering system with access to menu mix data will be able to quickly generate a suggested order that takes into account the forecast and quantities on hand.
- An integrated labor scheduling tool may break down expected sales or guest-counts by day-part, hour, or even fractions of an hour, making it simpler to schedule precisely for expected demand.
- The prep and production schedules can be mapped out for the period, minimizing the chances that food shortages or waste will occur.
Better yet, if the forecasts change, in an integrated system those changes will automatically flow through to the other modules, making on-the-fly adjustments (and the savings they represent) possible. In short, it’s integration with associated systems that makes extraordinary cost reductions possible with a forecasting system.
Advanced systems integrated with production modules may accommodate information as granular as hold times, thaw times, and shelf life of products.
3. Ease of Use
Nobody knows a restaurant like the person running it, so she should be able to create the forecasts in the field, instead of having to rely on someone at the corporate office to do so. That means the forecasting tool’s ease of use is key. There are actually two benefits to ease of use:
The forecast is more likely to be done using the forecasting tool, increasing accuracy and reducing waste and/or customer dissatisfaction.
If the manager isn’t present due to illness, vacation, etc. then it’s a simple matter to have someone else make the forecast using the tool.
In effect, an easy to use forecasting tool gives the restaurant a planning expert that’s always on the job– a highly desirable situation.
4. Level of Detail
Systems vary in how detailed a forecast the user can generate. For example, some systems will generate not just weekly, daily, and hourly sales forecasts, but may allow the user to specify 30- or even 15-minute increments. That can be enormously helpful in settings where traffic tends to come in concentrated bunches, for example in a restaurant with a heavy lunch rush.
In addition, systems that are integrated with production modules may accommodate information as granular as hold times, thaw times, and shelf life of products. The result is just-in-time production that reduces waste to an absolute minimum, while still meeting expected demand.
Forecasting is simple, but it’s not easy. It requires serious number crunching to produce an accurate model of future sales- but number crunching alone isn’t enough. To plan for the best possible guest experience and wring the maximum costs out of the operation, a forecasting application has to be easy to use, detailed, integrated with other food and labor planning applications, and must accommodate the metrics most important to the operator. A powerful forecasting system literally provides the roadmap to ultimate efficiency.