Skip to content
Predict.ai

Fundamentals

Seasonality

Seasonality is a pattern that repeats on a fixed clock — weekly, yearly, intraday — because people, weather, and operations do. It is not a trend, and it is not a one-off spike.

Retail has a Wednesday lull and a December pile-up. Grids have an evening peak. Hospitals have a Monday clinic bulge. If your forecast cannot say which weekday it is, it will spend its life apologizing.

The period matters. Weekly seasonality wants a 7-day lag. Yearly seasonality wants more than one trip around the sun, or a borrowed shape from a related series. Multiple seasons can stack: hour-of-day sitting on day-of-week sitting on month.

Do not confuse a one-time event with a season. A viral tweet is not 'seasonality.' A promo you run every Black Friday is closer — if you will run it again, put it in the calendar, not in the residual.

Weekly demand — 4 seasons + a promo

Seasonal demand history and a 7-week forecastpromonow27 weeks of history+7 week horizon
The weekly shape repeats. Week 19 is a promotion, not a new season. After 'now', the path is a forecast with a widening band.

Read next

Related terms

Try it on your data.

Connect an outcome and its history. Predict.ai finds the drivers, races the models, and keeps the forecast live — in the workspace, over the API, and through agents.