Forecast bias
Forecast bias is the consistent tendency of a forecast to run above or below actual results, found by averaging the signed errors instead of their absolute sizes.
What it means
Accuracy ignores direction, but bias keeps it. If forecasts are over actuals most months, the bias is positive and stock builds up. If they are mostly under, the bias is negative and customers go short. Random errors cancel out over time, while bias does not.
Example
A paint dealer forecasts monthly sales against actuals for six months: over by 8, 12, 5, 10, 7 and 9 percent. The average is plus 8.5 percent. The forecast is consistently high, and the dealer has carried about ₹4 lakh of unsold stock as a result.
Records and checks
- Record the forecast and actual per item and month, with the signed difference.
- Verify the bias figure over at least several periods before acting on it.
- The planner calculates it, and the sales head reviews whether targets or optimism are inflating forecasts.
Common mistakes
- Averaging absolute errors only, which hides a steady lean in one direction.
- Correcting for bias in the forecast while leaving the cause, such as sales targets, unchanged.
Did not find your answer?Message the lamp team on WhatsApp. Ask about setup, your records, or anything we have not covered here.Ask us on WhatsApp →