Market Research Project

[tina83]

Hello All,
I am a masters student and have decided to do a project on customer segmentation in SAS. The project is to identify the best type of customer companies which
might be interested in buying products from a cosmetic manufacturing company and the
type of deciding variables are the 1)Age of the potential buyer company 2) The turnover 3) Revenues 4) Whether it is a cosmetic company or not
These are just a few of the variables among the many others which is in the list.
I had these doubts here however:
1) How do we interpret the clusters once it comes in the form of a tree or otherwise?

2) How do we set these criteria in the programming part while working on SAS...I believe it is proc cluster or fastclus to be used but these simply give an output based on the Euclidean distance and give an output which has similar observations in the same cluster. However, if we were to set a specific "Criteria Variables" as mentioned( the above 4 variables) , how do we do the same in SAS...
Is there any other special procedure, statement or function used in SAS other than the proc fastclus/cluster
I tried to research a lot about this on the internet but didnt get much idea and neither any sample codes
I would really appreciate if I could get some inputs regarding this.
Thanking You,
With Sincere Regards,
Tina

[romakanta]

i know it's pretty late for a reply but just in case....

1. you will find lots of documentation on deciding the number of clusters but as far as interpretation is concerned, it's pretty much subjective just like the interpretaion of a factor analysis result. a lot will depend on the specific business/domain for which you are analyzing the data, and the business requirement of your client.

2. you can use PROC CLUSTER. specify the variables with VAR, and if you want to use a non-Euclidean distance for clustering, you can compute a distance matrix using PROC DISTANCE. you'll find details on the SAS online help doc.

i also guess your dataset will have mixed variable types. there are a lot of opinions on which clustering method should be used for such data but i find SPSS's two-step cluster the best as far as i know.

rgds,
romakanta