Forecast accuracy shows how close your demand forecasts are to what actually sold. Enter actual and forecast values for the same periods and the calculator returns MAPE, WAPE, bias, RMSE and the tracking signal.
Use it to compare forecasting methods, including AI models, against your current process on the same history.
Forecast Accuracy Calculator: MAPE, WAPE, Bias
Enter your numbers to see the result.
Estimates for planning only. Results depend on the accuracy of the inputs and the assumptions of the standard formula.
The formulas
For each period, error = actual – forecast. Over n periods:
- MAE (mean absolute error) = sum of |error| / n
- MAPE = average of |error| / actual x 100, skipping periods where actual is zero
- WAPE = sum of |error| / sum of actual x 100
- Bias = (sum of forecast – sum of actual) / sum of actual x 100; positive means over-forecasting
- RMSE = square root of the average squared error
- Tracking signal = sum of errors / MAE; values far from zero suggest persistent bias
WAPE is usually more reliable than MAPE when demand includes small or zero values, because MAPE is distorted by small denominators. Accuracy is often reported as 100% minus WAPE.
Worked example
Actual: 100, 120, 90, 110, 130, 95. Forecast: 105, 110, 100, 115, 120, 100.
- Errors (actual – forecast): -5, 10, -10, -5, 10, -5. Absolute errors total 45, so MAE = 7.5
- WAPE = 45 / 645 = 6.98%, so accuracy = 93.02%
- MAPE = average of 5.00%, 8.33%, 11.11%, 4.55%, 7.69% and 5.26% = 6.99%
- Bias = (650 – 645) / 645 = +0.78%, a slight over-forecast
- RMSE = square root of 62.5 = 7.91; tracking signal = -5 / 7.5 = -0.67
How to use the measures
Look at accuracy and bias together. A forecast can have low error on average but lean the same way every period, which causes steady overstock or stockouts. A tracking signal that keeps growing in one direction points to a bias to correct.
Measure at the level where decisions are made, usually item and location, and over a horizon that matches the lead time. Compare a new method with a simple baseline such as last period or a moving average on identical history. If the gain is small, keep the simpler method. See our demand planning guides and the safety stock calculator, which uses forecast variability.
Common questions
What is a good forecast accuracy?
It depends on the product and how stable demand is. Steady, high-volume items can be forecast much more accurately than intermittent or promotional ones. Track your own trend and compare against a simple baseline rather than chasing a fixed target.
Should I use MAPE or WAPE?
WAPE weights errors by volume and handles zero or small values better, so it is often preferred for planning. MAPE treats every period equally, which can overstate error on low-volume periods and breaks when actual demand is zero.
What does a positive bias mean?
The forecast is higher than actual demand on average, which tends to cause excess stock. A negative bias means under-forecasting and a higher risk of stockouts. This calculator shows over-forecast as a positive percentage and labels the direction in the result.
How many periods do I need?
The tool needs at least two, but more is better. A dozen or more periods gives a steadier picture, and the tracking signal in particular is unreliable on short samples, where one unusual period can dominate.