2011 Forecasting Performance Benchmark Study
Glossary
| Bias | Bias is calculated as the sum of signed errors (differences between forecast and shipments for each time period and item/location) divided by the sum of shipments. Positive and negative bias represents over-forecasting and under-forecasting, respectively. |
| Demand Planning (DP) |
The Demand Planning solutions employed by participating companies to create forecasts based on the time-series analysis of historical data and augmented to reflect promotions and planner insight/value-add. |
| Demand Sensing (DS) |
Terra Technology’s Demand Sensing solution that leverages new mathematical techniques and near real-time information to create a more accurate forecast of demand, based on the current realities of the supply chain. |
| Global Velocity | Item base codes of the entire dataset ordered and parsed into five equal groups by volume. Volume 1 refers to the fastest moving products. |
| Horizon | Horizon is the cumulative length of time in days starting from the forecast creation date. Horizon 7 refers to the current week and is analogous to weekly Lag 1. |
| Lag | Lag refers to the number of weekly or monthly buckets from forecast creation date to the forecast date. |
| Local Velocity | Item base codes of the company-specific dataset ordered and parsed into five equal groups by volume for each company. Volume 1 refers to the fastest moving products. |
| Mean Absolute Percentage Error (MAPE) |
MAPE expresses forecast accuracy as a percentage and is calculated as the sum of absolute errors (differences between forecast and shipments for each time period and item/location) divided by the sum of shipments. MAPE is always positive. |
| Naïve Forecast (Nv) |
A simplistic forecast based on a seasonally-adjusted moving average. The naïve forecast provides a means to compare the relative value of sophisticated forecasting tools such as Demand Planning and Demand Sensing. |
| Naïve versus Demand Planning (Nv vs. DP) |
The relative comparison between the Naïve and Demand Planning forecast errors. This calculation divides the difference between the naive MAPE and DP MAPE by DP MAPE. |
| Version | Forecast creation date. |
