Supply Chain Management Blogs by SAP
Expand your SAP SCM knowledge and stay informed about supply chain management technology and solutions with blog posts by SAP. Follow and stay connected.
Showing results for 
Search instead for 
Did you mean: 
One of the things we in SAP Solution Engineering explored with our customers is the usage of 'Like Modelling' and 'Phase-in' curves. The blog below shares our experience in when to use each of these techniques while forecasting new products.

In like modelling, when we borrow the history from 'Reference Product(s)', we are mainly interested in the growth patterns (trend). However, the history contains trend + seasonality+ events + noise. We try to minimize the noise by using a group of products with varying weights instead of a single product. However, the history might still contain events and seasonality. Let us say, the borrowed history pertains the products (say juices) that were introduced before the summer season. Then the history would have not only the growth pattern but the seasonal impact as well. If the new product is being introduced in a different season, it is important to neutralize the seasonal impact in the history.
This is where, the ramp up profiles offer an advantage over the like modelling. They provide different kinds of ramp-up profiles. A sub-linear ramp-up profile of quadratic/cubic type provides a slow growth initially and scales up faster later on. A super linear growth provides high initial growth tapering off later. Based on the previous launch experiences the users can take these profiles as the starting point and then apply any other effects like a seasonal impact or a promotion. This helps us avoid the kind of over fitting we do with 'Like Modelling'.

Choice of Sub-linear/Super-linear: When products are launched, there would be innovators who like experimenting and would be the early purchasers. Then you have the imitators who buy the product based on the experiences of the 'experimenters'.Initial growth can be fast/slow (sub-linear/super-linear) based on the type of the product, the proportions of experimenters/imitators in the target segment and the pre-launch/early-launch marketing efforts. Forecasters broadly know this nature of the customer segments based on the previous launch experiences. Thus they can choose the relevant characteristic curve as the starting point and apply these growth patterns on the stable level volume that they would have assessed for the product.

This flexible new product forecasting method together with product substitution provides superior capabilities of demand planning & channel filling in IBP while launching new products.