Applying causal factors to demand

When forecasting demand, you can apply causative data to your demand data and examine the effects of these factors on your demand model. Causative data can consist of various external factors such as weather and economic conditions, as well as internal activities such as product promotions and other site activities.

To do this, you create causal time series definitions to match your demand time series definitions, and then apply these causals during demand model analysis. Each causal time series definition can include multiple time series, from a source that is one of the following:

• user-defined
• based on promotional data, to be applied at the definition level
• input via Trend Cloud

Once created, you can:

• Select one or more definitions as input when creating or editing a model.
• Apply the definition during demand model analysis.

You can also edit a causal time series definition as needed.

The use of causal data is not required to model demand, but it typically results in more accurate forecasts because you applying real-world conditions that can affect your model.

For instructions on defining causal time series for your demand time series, see Create a causal time series definition