Learn · forecast accuracy metrics

    Forecast Accuracy Metrics Explained

    A forecast without an accuracy record is an opinion with decimal places. Forecast accuracy metrics report how far past forecasts missed, in which direction, and how consistently — which is what tells a team how much weight the next forecast deserves.

    Error size: MAE and RMSE

    Mean absolute error is the average miss in the same units as the forecast — 40 units, 3,000 dollars. Root mean squared error is similar but penalizes large misses more, so it is the better guide when one bad month hurts far more than several small ones.

    Error as a percentage: MAPE

    Mean absolute percentage error expresses the miss relative to actual volume, which makes forecasts for different products or regions comparable. It becomes unstable when actual values approach zero, so it is reported alongside an absolute measure rather than alone.

    Bias: consistently high or low

    A forecast can have a small average error and still be wrong in one direction every period. Bias measures that lean. Persistent bias usually points to a process problem — a target treated as a forecast — rather than a modeling problem.

    Comparing against a simple baseline

    A forecast should be judged against the naive alternative of repeating last period or last year. A sophisticated model that does not beat that baseline is not adding value, and the comparison is the honest test.

    How the results are reported

    Averisys time series analysis reports forecast accuracy in plain English, states how much confidence the numbers deserve, and flags where the forecast is least reliable. A free trial is available without a credit card.

    See Averisys time series analysis

    Explore the forecast accuracy metrics from Averisys Analytics. Watch a demo and start a free trial — no credit card required.

    Visit Averisys time series analysis

    Frequently asked questions

    What are forecast accuracy metrics?

    Measures of how far forecasts differ from actual results, used to judge how much a forecast can be trusted.

    What is MAPE?

    Mean absolute percentage error: the average miss expressed as a percentage of the actual value.

    Which metric should be used?

    Report an absolute measure such as MAE with a percentage measure and a bias measure; no single number is sufficient.

    What does forecast bias mean?

    That forecasts are consistently too high or too low, rather than missing in both directions.

    What counts as good accuracy?

    It depends on the series; the practical test is whether the forecast beats simply repeating the last period or last year.

    Related guides