Inventory glossary

What is Exponential smoothing?

Exponential smoothing is a way to forecast future sales by weighting recent sales history more heavily than older history, so the forecast reacts to recent changes without overreacting to one unusual day.

For a small Shopify store, sales rarely sit still. A candle store might sell 10 units a day for weeks then jump to 25 a day after a social media mention. Exponential smoothing gives you a forecast that adjusts toward that new level gradually, instead of treating one big spike as the new normal or ignoring it completely.

In practice you use it to set a baseline daily or weekly sales number per SKU, which then feeds into reorder points and order quantities. You do not need to run the math by hand. The useful part is understanding that the forecast gives more weight to this week than to last month, and that you can control how fast it reacts by adjusting the smoothing factor.

Formula

Forecast(t) = alpha x Actual(t-1) + (1 - alpha) x Forecast(t-1), where alpha is between 0 and 1

Example

A coffee roaster sells 40 bags of a blend last week and the smoothed forecast going into that week was 35 bags. Using alpha = 0.3: new forecast = 0.3 x 40 + 0.7 x 35 = 12 + 24.5 = 36.5 bags. The forecast moves up from 35 toward 40, but not all the way, so one strong week does not fully reset expectations.

Common mistakes

  • Picking an alpha that is too high, which makes the forecast swing wildly after every unusual sales day.
  • Using it on a brand new SKU with no sales history, which gives a meaningless starting forecast.
  • Treating the smoothed forecast as exact and ignoring known future events like a planned promotion or holiday.

Exponential smoothing and SteadyShelf

SteadyShelf uses this kind of smoothed sales pattern per variant to calculate an order-by date and suggested order quantity for each SKU, so you do not have to track the math yourself.

SteadyShelf is launching on the Shopify App Store. Join early access to get the install link first.

Related terms

Demand forecasting · Moving average forecast · Weighted moving average · Forecast error · Seasonal forecasting

Full inventory glossary

FAQ

Does exponential smoothing work for seasonal products?

Basic exponential smoothing does not account for seasonality on its own. For products with clear seasonal swings, like holiday-themed items, you need a seasonal version of the method or a separate seasonal adjustment, otherwise the forecast will lag behind real demand changes.

What alpha value should a small store use?

There is no universal number. A lower alpha, such as 0.1 to 0.2, gives a steadier forecast that is good for stable products. A higher alpha, such as 0.4 to 0.6, reacts faster and suits products with frequent demand shifts. Testing against your own past sales is the only reliable way to pick one.

Let SteadyShelf do this math for every SKU

Order-by dates per supplier, seasonal forecasts and draft POs for your Shopify store. $29/month after a 30-day free trial.

Get early access