Demand forecasting
Exponential Smoothing
By the Supply Chain Manager team · Updated October 16, 2026
Exponential smoothing forecasts demand as a weighted average of past demand in which the newest data counts most and older data fades gradually.
A smoothing factor, alpha (between 0 and 1), sets how quickly the forecast reacts. Simple exponential smoothing tracks a level. Holt's method adds a trend, and Holt-Winters adds a repeating seasonal pattern on top.
How to calculate exponential smoothing
New Forecast = α × Latest Actual + (1 − α) × Previous Forecast
Exponential smoothing example
Illustrative numbers: last forecast 100, latest actual 120, α = 0.2. New forecast = 0.2 × 120 + 0.8 × 100 = 104. With α = 0.5 it would jump to 110, reacting faster but also chasing more noise.
Why exponential smoothing matters for distributors
Exponential smoothing methods are cheap to compute and adapt well, which suits a distributor forecasting thousands of SKUs every night.
Common exponential smoothing mistakes
- Picking alpha by feel instead of testing it on history.
- Using simple smoothing on strongly seasonal SKUs, where it will always lag the season.
- Applying it to intermittent demand, where Croston's method usually does better.
Exponential smoothing in Supply Chain Manager
Supply Chain Manager backtests simple exponential smoothing, Holt's linear trend and Holt-Winters with a weekly cycle for each SKU.
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.