SAP Integrated Business Planning (IBP) Forecast Automation
Do you agree writing a blog post is not that easy because every individual has a different perspective? Sometimes after publishing the post need to check if any points to be changed or added based on comments. I usually do modifications based on responses from various groups and experts.
I was more interested to learn as much as product information and present it in a blog post to help others how the functionality and feature will work in an SAP environment. I do not know suddenly my thought process changed rather focusing on product information and how the feature and functionality will help from a business user’s perspective. In the end business users live with the features that were implemented during the implementation.
I would like to more focus on forecast automation in this post, in the earlier post Forecast Algorithms and their importance was covered. The forecast automation term seems to be fancy and easy but achieving automation is not an easy process. SAP IBP forecast automation supports planners to understand the patterns of forecast data i,e tie series data, it helps to take the right decisions.
The main goal of forecast automation is to reduce the efforts to understand the forecast data available in the application. I am interested to add my experience with the forecast automation process to one of the fortune 500 companies.
The project scope was to implement S&OP and Demand process, as usual, we started with requirements gathering and business process mappings. Everything going on good during the user acceptance test one of the users raised a concern that the Forecast Automation process is not covered in the scope. As you know everyone started focusing on automation rather than UAT. We convinced the team to continue with UAT we will add this process to the product backlog and deliver it as part of the project scope.
We were ready with a forecast automation demo before we start the discussion users want the system to demand to forecast automatically. Demand planners will review the results and come to a consensus plan. In short, they do not want to manually select the forecast algorithm system should select based on sales history.
We started with the demo it went well overall, after the demo one of the user said it is very simple and we can configure it ourselves and test it accordingly. Imagine in this situation every IBP consultant is happy because the business users themselves explore however consultant help is required to set up initial configuration settings.
In order to achieve forecast automation in IBP following sub-processes are available.
- Time series analysis
- Change point detection
We will see how these two sub-processes work in real-time scenarios.
Time series analysis will identify the sales history patterns and segregate them as trend, seasonality, and intermittency. These patterns were already covered in an earlier post and nothing new here.
|Continuous||Time series with very few missing values and no other patterns|
|Intermittent||Time series with a large number of zero values|
|Volatility||Intermittent time series in which relatively large deviations can be observed|
|Irregular||Continuous time series with high volatility but no specific pattern|
Change point detection will analyze the data points and their changes in sales history.
Ok, so far so good to understand the time series and change point detection but how to model this feature end to end?
The answer follows as,
- Creation of forecast automation profile
- Run the application job template
- Analyze the results of time series and change point detection
I would recommend watching the below video on how to set up Forecast Automation in SAP IBP.
The forecast automation feature saves demand planners time and resources to achieve an accurate forecast for future horizons.
Sometimes we may not know everything but we have the opportunity to learn and support businesses to plan better.
I thank you for reading this blog and soon will catch up with another interesting topic.