Data is the foundation of a forecasting process. Making sure that your data is well organized, accurate and appropriately set up is a critical first step when you are setting up your forecasting process.
Forecast Pro works with historic time series, i.e., sequences of values of a variable at some equally spaced time interval. Business series are usually expressed in terms of the calendar, and can be daily, weekly, monthly or quarterly. Forecast Pro can work with any type of time series but is oriented towards weekly and monthly calendar data.
There are several key questions you will want to think about when setting up your data, including:
How much data history should I use?
What data format should I use?
Do I want to connect with ODBC or use Excel/text files?
Do I need to interface with an ERP system?
How should I set up my hierarchy to generate the most accurate forecasts?
What data attributes should I include in my data files?
What data do I want to use in addition to historic demand data? Do I need conversions, filters or external data?
Is there additional data I will need for customized forecasting models?
When you are first getting started with Forecast Pro, you should consult with your Forecast Pro trainer and your IT department to help you answer these questions. You should also read the following key chapters in The Forecast Pro Reference Manual: Setting Up Your Historic Data, Setting up your Optional Data and Working with Hierarchies. Once you set up your data hierarchy and attributes, changing the data structure may necessitate losing your forecast archive and starting from scratch, so you should try to get it right from the beginning.
The purpose of this lesson is not to answer all of the above questions, but rather to provide an interactive introduction to some of the data concepts in Forecast Pro and explain some of the connections between your data files and the settings in Forecast Pro. To do so, we will describe how to set up your historic data in Excel using Forecast Pro’s row format.
If you are using a Collaborator license: reading in historic data is not supported. Please proceed to the next lesson.