How can you improve delivery costs per gallon?
Propane Delivery Company
The company was looking for significant cost savings across the board, and last-mile deliveries carried high costs and the most critical risk to the business’s health. In the fuel delivery business, refilling a customer early is marginally more expensive; getting to a customer late means they are out of fuel and calling your competitors. The goal is to balance under-servicing (losing customers) and over-servicing (losing money).
In the fuel delivery business, there are three key metrics:
- Out of Gas – The customer is bone dry on refill.
- Zero Gallon Delivery – You deliver, but there is no room in the tank.
- Early Delivery – Customer on the route that doesn’t need a delivery.
Early deliveries are by far the most common. An 8-hour shift delivering propane costs the same whether you provide 1,000 gallons or $2,500 gallons, although the cost to deliver the propane is significantly higher in the former. Therefore, the goal was to determine the optimal number of drivers to minimize early deliveries.
To quantify the opportunity, we created a monthly driver efficiency score model based on the number of gallons remaining in the truck at the end of the day. A driver efficiency score of 1.0 meant you finished your shift with an empty truck. A driver efficiency score of 2.0 meant you finished your work shift with a truck that was half full. Your delivery expense for both routes was equal. The model allowed us to identify over and under-capacity based on location and seasonality to augment workflow appropriately.
- MS SQL
- Data Mining
- Field Visits
- Driver delivery ride alongs
- Excel modeling
The amount of propane sold over a year will be the same regardless of delivery schedule, absent period timing differences that may fluctuate over month end. The only certainty in this business is that minimizing early deliveries with running customers out of gas significantly increases profit margins.
Fuel route planning is very complicated, and factors that contribute to early deliveries are typically the weather in general and the change of season in particular. Early deliveries were 78%, but out-of-gas deliveries were only 0.4%. The theory was that the squeaky wheel not being quantified contributed to the number of premature deliveries. A 50% reduction in these deliveries was worth approximately $30 million.