Forecast Accuracy and MAPE: A Practical Primer
Forecast accuracy explained: MAPE, WAPE, bias, and how to use them to improve retail forecasting.
Forecast next-period demand using a weighted moving average of three recent periods.
Next-Period Forecast
537.0 units
Simple Average Comparison
510.0 units
Recent Trend
+42.9%
Period -1 (most recent)
600 units
Formula Used
(0.2 × P-3) + (0.3 × P-2) + (0.5 × P-1)
(0.2 × P-3) + (0.3 × P-2) + (0.5 × P-1)
Weighted moving average gives more weight to recent periods. Default weights 0.2 / 0.3 / 0.5 emphasize the latest trend without overreacting to a single spike.
Recent sales: 420, 510, 600 units. Forecast = (0.2 × 420) + (0.3 × 510) + (0.5 × 600) = 84 + 153 + 300 = 537 units. Use this as the baseline forecast for the next period.
For SKUs with strong seasonality or promotional volatility. Use seasonal naïve, ETS, or ML models instead.
Yes — pick weights that sum to 1.0 and reflect how much you trust each recent period.
Deep-dive guides that explain the math behind this calculator.
Forecast accuracy explained: MAPE, WAPE, bias, and how to use them to improve retail forecasting.
Demand sensing in retail: short-term signals, real-time data, and how to act faster than traditional forecasting allows.