Stockertim Dictionary · Strategy and management · Technology
Exponential smoothing
Also called simple exponential smoothing, single exponential smoothing
In short
Exponential smoothing is a forecasting method that adjusts the previous forecast by a fraction of the last forecast error, so recent demand counts more than older demand.
- Intermediate
- Ratio or metric
- Checked
In detail
Exponential smoothing produces each new forecast from two numbers: the last forecast and the last actual result. The new forecast moves part of the way from the old forecast towards what actually happened. How far it moves is set by a smoothing constant, usually called alpha, between 0 and 1. Because each forecast includes the one before it, every past observation still has an influence, but that influence shrinks geometrically with age.
The choice of alpha balances responsiveness against stability. A high alpha, such as 0.5, reacts quickly to changes in demand but also chases random noise. A low alpha, such as 0.1, gives a smooth forecast that ignores one-off spikes but is slow to notice a genuine shift. Analysts often choose alpha by testing which value would have given the smallest errors on past data.
Simple exponential smoothing suits series with no clear trend or seasonal pattern. Extensions add a trend term, known as double or Holt's method, and a seasonal term, known as triple or Holt-Winters smoothing. The method is popular because it needs little data storage, is easy to explain and performs well for many short-term demand forecasts.
Formula
New forecast = old forecast + alpha × (actual − old forecast)
- old forecast
- the forecast made for the period just ended
- actual
- the demand that actually occurred
- alpha
- the smoothing constant between 0 and 1
Worked example
A café chain, Brew Lane, forecast 500 coffees for Monday and sold 540. It uses a smoothing constant of 0.2.
- Forecast error: 540 − 500 = 40
- Adjustment: 0.2 × 40 = 8
- Next forecast: 500 + 8 = 508
Brew Lane's forecast for the next Monday is 508 coffees.
In practice
Inventory and demand planning systems use exponential smoothing for thousands of items at once, and statistical handbooks describe how to choose the smoothing constant from past errors.
Watch out
- Simple smoothing lags behind a steady trend; use a trend-adjusted version.
- A forecast that looks smooth may still be consistently biased.
Learn it properly
Official sources
- 6.4.3 What is exponential smoothing? (NIST/SEMATECH e-Handbook) · National Institute of Standards and Technology
- 6.4.3.1 Single exponential smoothing (NIST/SEMATECH e-Handbook) · National Institute of Standards and Technology
- 6.4.3.2 Forecasting with single exponential smoothing (NIST/SEMATECH e-Handbook) · National Institute of Standards and Technology
Facts last checked on 3 November 2026.
Education, not advice: the Stockertim Dictionary explains what terms mean. It never recommends buying or selling anything.