Seasonality adjustment

Seasonality adjustment statistically estimates and removes recurring calendar effects from an original series to make underlying developments easier to assess.[1] Seasonal effects are not limited to seasons of the year: a fixed day-of-week pattern can matter too. Seasonal patterns have a fixed, usually known calendar period, unlike cyclical fluctuations without a fixed period.[2] Not every temporary increase in sales is therefore a seasonal effect.

For a store, model-based adjustment could help investigate growth beyond a recurring gift-shopping peak.[3] Comparing equivalent seasonal periods is a useful first check, but does not by itself create a formally seasonally adjusted series.[1] Retain raw figures alongside adjusted results and document the method. Track one-off promotions, stockouts and moving holidays separately so genuine changes are not casually explained away as seasonality.