Twin Peaks Implements Predictive Ordering to Save Millions
The Solution
Decision Logic’s Predictive Ordering feature provides suggested order quantities to managers by analyzing your prior sales and usage data. There are several prediction methods considered when calculating the suggested order quantities:
- Actual Usage
- Ideal Usage
- Total Sales
- Bar Sales
- Food Sales
- Reminders
These methods are applied to each ingredient so you are not tied to one method per distributor. Once deployed, managers are provided with suggested quantities when they place an electronic order through Decision Logic. Our analysis automatically takes into consideration any anomalies that have been detected in the prior Sales and Usage data. The manager can also make adjustments to the suggested order based on known variables not accounted for in the past for example, a new event this week. The Predictive Ordering feature significantly reduces the amount of time it takes to place an order and optimizes the quality of decision making around the order. The result is the correct amount of food on your shelves, therefore freeing up cash flow and decreasing waste.
Since implementation, Twin Peaks stores are ordering on average $2,000 less on food per month, with increased sales of $5,000-10,000. Talk about right sized inventory!
Case Study Example
The AFG team was preparing 25 cases of tomatoes per store, per week, for all of the recipes that use fresh tomatoes. The core functionality of Decision Logic was showing they had a large tomato waste. Utilizing Production Prep Sheet, they were alerted to the proper amount of preparation to minimize waste. The features work together to showed AFG operational improvements that were needed. In addition, AFG began to investigate operationally and found that recipes were not being followed. Visually, 6 tablespoons of diced tomatoes did not appear to be much different than 2 tablespoons on a plate of nachos. They are now prepping 40% less of the tomatoes per week and dropping those savings to the bottom line.