How to estimate your number of predictions?

PlanIQ pricing is based on number of predictions per month. Let’s define what those are and how to estimate them.

 

A prediction is made when “Forecast Actions” are executed. Each Item that is part of a Forecast Action execution output is considered a prediction for this purpose. For example, in a Supply Chain use case, this might be a SKU or a SKU-store combination. For a Finance use case, this might be a budget forecast line item, for example spend category or cost center. For Workforce planning, a prediction would be a resource that is impacting the demand, either directly, for example call-center employee utilization, or indirectly, for example number of customer calls or support tickets.

 

Demand Planning Example:

  • 1000 SKUs across 3 countries, each with 100 stores
  • Your number of predictions per execution of “Forecast Action” would then be 1000 x 3 x 100 = 300,000*

 

If you execute single forecast operation per month, then those 300,000 is what you need.

If you execute 4 weekly forecasts per month, you will need 1,200,000 monthly predictions.

 

If you have multiple forecast actions running against multiple forecast models it would incur higher prediction consumption.

 

PlanIQ pricing or consumption does not change depending on the length of the forecast horizon or the granularity of data.

 

*Note: this calculation assumes that all SKUs are equally available in all locations. Your actual scenario might be different.

0 Kudos
Version history
Last update:
‎07-07-2021 12:18 PM
Updated by:
About the Author
Contributors
Labels (2)