Forecast dataset

Forecast dataset containing demand forecast values per product and location, an optional forward-looking demand signal for Inventory Management.

The forecast dataset contains demand forecast values for each product at each location. It provides an optional forward-looking demand signal that customers can supply if they already produce their own forecasts.

Purpose

The inventory management solution uses forecast data as an additional demand signal to complement historical demand analysis. When customers already generate their own forecasts, this dataset allows those forward-looking values to inform replenishment and stock optimization recommendations.

Required fields

About the Constraint column: every field listed must be provided in your data. Non-null means its value cannot be null; Nullable means the value may be sent as null (or left blank in your source). Primary key, Unique, and Foreign key describe the field's role in the data model; a Primary key is always non-null.

Field	Description	Type	Constraint
product_id	Unique identifier for the product.	string	Primary key
location_id	Unique identifier for the store or distribution centre.	string	Primary key
cutoff_date	The date on which the forecast was generated.	date	Primary key
start_date	First date of the forecast period.	date	Primary key
end_date	Last date of the forecast period.	date	Non-null
horizon	Number of periods ahead of the cutoff date that the forecast covers.	integer	Non-null
yhat	Forecasted demand value for the period.	float	Non-null
frequency	Length of the forecast period, in days.	integer	Non-null

Usage notes

• This dataset is optional and only required when customers already produce their own demand forecasts.
• Product and location identifiers must align with the corresponding Products and Locations datasets.
• Forecast periods should be consistent, with frequency defining the length of each period in days.
• Forecast values should reflect expected demand, not historical actuals.