Like modeling basics

[Former Member]

We have new products being introduced in our DP project and it is primarily one to one replacement, there are 100 new products being used.

1.Could some one let me know if in this scenario we should be using like modeling (in this case as this is one to one replacement 100 like modeling profiles would be required) or should we be using Realignment by which we can mass generate the CVC and then just copy the history from one CVC to another.

2.What are the situations in which like modeling is going to be used as compared to realignment.

Thanks in advance

[Former Member · Accepted solution]

Hi.

If it is true 1 for 1 replacement, then realignment would be the easiest route.

Like modelling allows you to use other materials' history as a starting point which you can then model together with other products and profile how this history applies to the new material. e.g, New product history using 40% Product A, 60% Product B and then 120% of the combined sum.

You can also incorporate Phase In and Phase Out profiles to manage transitions from old to new products.

If these functions are not required, then realignment works just fine.

Regards, Mark

[Former Member · Accepted solution]

Do you know of a list that where we can see all the reference characteristic', like model and the new product and the weight age profile. I have been able to a list out where we have all the above mentioned things but not the reference characteristic values,

Do u have an idea about this

[Former Member]

HI APO APO.

The reference characteristic is determined by the Forecast profile. In the Forecast Profile that applies to the characteristic combination in question, the historical key figure is determined. Therefore, there is not necessarily a 1:1 relationship between the reference characteristic and the like profile. I think you already understand this judging by the grouping of data you are mentioning. I can only suggest qurying with a join on the assigned forecast profile (if such an assignment exists, which will depend on your approach to forecasting).

Hope this helps, Mark.