FORECAST QUALITY & MEASUREMENT
Forecast Error
Forecast error is the difference between a forecast and the actual result for a matched item, location, period and unit. The sign convention and aggregation method must be stated.
FORECAST ERROR IN PRACTICE
- Match the recordsUse the forecast version available when planning.
- State the calculationDefine sign, units and metric.
- Interpret the patternReview bias and error size separately.
Use the agreed units, actual records and approved procedures for your operation.
A WORKED EXAMPLE
−20 and +20 cancel, but both miss.
Using actual minus forecast, a forecast of 100 with actual 80 gives −20; another forecast of 100 with actual 120 gives +20. Mean signed error is zero, while mean absolute error is 20. The forecast was not perfect.
WHAT THIS MEANS FOR YOUR QUOTE
Agree the requirement before comparing quotes
Use the actual shipment, stock and service requirements. Record the billing unit, included work and responsibility for exceptions.
A LITTLE MORE CONTEXT
Questions about Forecast Error
Does zero bias mean perfect accuracy?
No. Positive and negative errors can cancel.
Is the sign convention universal?
No. State it.
Can percentage error fail at zero actual?
Yes. Define an appropriate treatment.
Source & example notes
Definition and operational example prepared by 3PL Hub. Examples describe hypothetical requirements; verify the actual records and agreed procedures for your operation.
Quantities, costs and scenarios are illustrative, not operating targets, approved handling instructions or guarantees. Source guidance reviewed 22 October 2026.
Last updated: · 3PL Hub
