Supply Chain Management (SCM) 101: Demand Forecasting Forecast Errors Bias MAD MSE MAPE

What This Is

Forecast errors are a critical aspect of supply chain management, as they can lead to stockouts, overstocking, and wasted resources. A forecast error occurs when the actual demand differs from the predicted demand. For instance, Amazon, a company known for its efficient supply chain, uses advanced forecasting techniques to predict demand for its products. If Amazon underestimates demand for a popular product, it may lead to stockouts, resulting in lost sales and customer dissatisfaction.

Key Frameworks & Formulas

• Bias: A measure of the difference between the forecast and the actual value, indicating whether the forecast is consistently too high or too low. Bias = (Σ(actual - forecast)) / n
• MAD (Mean Absolute Deviation): A measure of the average difference between the forecast and the actual value, indicating the accuracy of the forecast. MAD = Σ|actual - forecast| / n
• MSE (Mean Squared Error): A measure of the average squared difference between the forecast and the actual value, indicating the accuracy of the forecast. MSE = Σ(actual - forecast)^2 / n
• MAPE (Mean Absolute Percentage Error): A measure of the average percentage difference between the forecast and the actual value, indicating the accuracy of the forecast. MAPE = Σ|actual - forecast| / (Σactual) * 100
• Safety Stock: The additional inventory held to mitigate the risk of stockouts due to forecast errors. Safety Stock = Z × σ × √L
• Lead Time: The time it takes for inventory to be replenished from the supplier. Lead Time = 5 days
• Service Level: The percentage of demand that is met from stock on hand. Service Level = 95%
• Fisher's Model: A framework for classifying products based on their demand variability and lead time. Fisher's Model = (Demand Variability, Lead Time)
• SCOR (Supply Chain Operations Reference): A framework for evaluating and improving supply chain performance. SCOR = (Plan, Source, Make, Deliver, Return)

Step-by-Step Application

1. Calculate the safety stock required for a product with a demand of 100 units per day, a lead time of 5 days, a standard deviation of 10 units, and a service level of 95%. Safety Stock = Z × σ × √L = 1.645 × 10 × √5 = 45.4 units
2. Determine the reorder point for a product with a demand of 100 units per day, a lead time of 5 days, and a safety stock of 45.4 units. Reorder Point = Demand × Lead Time = 100 × 5 = 500 units
3. Evaluate the performance of a supply chain using the SCOR framework. SCOR = (Plan, Source, Make, Deliver, Return) = (80, 90, 85, 95, 90)

Common Mistakes

• Mistake: Assuming that the forecast error is always due to the demand variability.
  • Correction: The forecast error can be due to various factors, including demand variability, lead time, and forecasting errors.
• Mistake: Using the MAD as the only measure of forecast accuracy.
  • Correction: The MAPE and MSE should also be used to evaluate the forecast accuracy.
• Mistake: Ignoring the safety stock when calculating the reorder point.
  • Correction: The safety stock should be included when calculating the reorder point to ensure that the product is always available.

Exam / Certification Tips

• Tip: Be able to calculate the safety stock and reorder point using the formulas.
• Tip: Understand the different types of forecast errors and how to mitigate them.
• Tip: Be able to evaluate the performance of a supply chain using the SCOR framework.

Quick Practice Problem

A company has a demand of 100 units per day for a product with a lead time of 5 days. The standard deviation of the demand is 10 units, and the service level is 95%. What is the safety stock required for this product?

Answer: 45.4 units

Explanation: Safety Stock = Z × σ × √L = 1.645 × 10 × √5 = 45.4 units

Last-Minute Cram Sheet

• Bias: A measure of the difference between the forecast and the actual value.
• MAD (Mean Absolute Deviation): A measure of the average difference between the forecast and the actual value.
• MSE (Mean Squared Error): A measure of the average squared difference between the forecast and the actual value.
• MAPE (Mean Absolute Percentage Error): A measure of the average percentage difference between the forecast and the actual value.
• Safety Stock: The additional inventory held to mitigate the risk of stockouts due to forecast errors.
• Lead Time: The time it takes for inventory to be replenished from the supplier.
• Service Level: The percentage of demand that is met from stock on hand.
• Fisher's Model: A framework for classifying products based on their demand variability and lead time.
• SCOR (Supply Chain Operations Reference): A framework for evaluating and improving supply chain performance.
• ⚠️ Reorder Point: The point at which the inventory should be replenished to meet the demand.
• ⚠️ Safety Stock: The additional inventory held to mitigate the risk of stockouts due to forecast errors.
• ⚠️ Lead Time: The time it takes for inventory to be replenished from the supplier.
• ⚠️ Service Level: The percentage of demand that is met from stock on hand.