Navigating tomorrow: Mastering the art of forecasting today

#1

This session will take place on Wednesday April 3rd at Prophix Live! 2024

image
image.png1276×586 103 KB

Let us know if you have any questions for Ujjwal down below!

#2

Hey Ujjwal, How can i use predictive forecasting to integrate my current short term data model and forecast m y financials

#3

Hi Sneha,
Predictive Forecasting is a great tool to help with forecasting your financials. It does work best with long term data as it learns the seasonality and trends in your data and applies it to your upcoming model.

In process manager you can use the Predictive Forecasting tool, it will ask you for the member combination of your training data and your destination. Once selected you should be able to run the process for a forecast. Check out the help article for this tool for best practices as well.

1 Like
#4

@ujjwal.sewani, how much historical data do you recommend for a decent predictive forecast?

#5

Hi @bob.smiley,

Seasonality in the data is based on the number of period in the source data. The process should ideally be run with at least one full years worth of data. If the total number of combinations are greater than 200,000 it wont remain efficient. To check for the number of combinations, you can multiply the number of selected members in each dimensions.

Hope that provides you an idea of the recommended minimum and maximum data sets.

1 Like
#6

Hello, Is it possible to overload the forecasting models with too much data? What is the max number of accounts and departments recommended>