Demand Segmentation and Forecastability Analysis

Are you creating Forecasting classes using your Supply chain’s ABC classification? Bad idea!!

I have preached Usability and Model tuning for demand planning applications like SAP, JDA Manu, Demantra, Logility etc.  A key stepping stone to Model tuning is the Demand Segmentation and Forecastability Analysis.

In our Usability consulting, we offer Demand Segmentation as part of the total package.  However, more and more clients are requesting us to explore their data and advise them on Demand Segmentation and creating Forecast classes that will involve their SKUs being bucketed in different segments.   Then we will run a data pattern analysis on the underlying data to determine what modeling algorithms may best match the forecast classes.

This is an important exercise that provides a roadmap for the client to start the usability process.  Since it involves a lot of data mining and offline analytics to determine this based on millions of rows of data, it does make sense for this exercise to be outsourced.  The client company may not have the data powerhouse or a software to do this.

Ideally this may need to be refreshed once a year but who will be responsible for doing this refresh?  If the client is large enough, they may have a Center of Excellence group to perform this.

To mitigate this, we are experimenting an offering that will take up your demand segmentation exercise.  If you want to work with us, we will do this on an outsourced basis in two weeks or less using the following steps:

  1. You provide the data in Excel or Access
  2. You answer a questionnaire that gives us info about the data
  3. Pay a nominal upfront fee
  4. We do the analysis
  5. Provide you the results and present the findings in a webinar or in-person presentation

Please contact http://demandplanning.net if you are interested in becoming an early adopter of this outsourced service at a very nominal cost.

Leave a Reply

Your email address will not be published. Required fields are marked *

*