Demand forecasting
MASE (Mean Absolute Scaled Error)
By the Supply Chain Manager team · Updated October 16, 2026
MASE (mean absolute scaled error) compares a forecast's average error with that of a simple naive forecast. Below 1 beats naive; above 1 is worse.
Scaling by the naive method's error makes MASE comparable across SKUs of any volume, and it still works when demand is zero in some periods. It answers a blunt question: is this method actually better than assuming tomorrow looks like today?
How to calculate MASE
MASE = MAE of the forecast ÷ MAE of a naive forecast on the same history MAE = mean of |Actual − Forecast|
MASE example
Illustrative numbers: a smoothing method misses by 6 units a week on average; repeating last week's sales misses by 8. MASE = 6 ÷ 8 = 0.75, so the method removes about a quarter of the naive error.
Why MASE matters for distributors
A forecast that cannot beat naive is adding complexity, not value. MASE makes that visible per SKU, so you know where to trust a model and where a simple rule will do.
Common MASE mistakes
- Reading a MASE of 0.9 as '90% accurate'. It is a ratio to a benchmark, not an accuracy percentage.
- Comparing MASE figures that used different naive benchmarks (for example last day versus same day last week).
MASE in Supply Chain Manager
Supply Chain Manager shows MASE on the Forecast Accuracy page, and a MASE above 1 is labelled as behind the naive benchmark.
Open the live demoExamples on this page use made-up round numbers to show the calculation. They are not benchmarks or results from real businesses.