Programming the statistical procedures from SAS

Customer Behavior Segmentation - 2 stage Clustering

Occasional Contributor Fae
Occasional Contributor
Posts: 14

Customer Behavior Segmentation - 2 stage Clustering

[ Edited ]

I am trying to do customer behavior segmentation using Two stage clustering - Fastclus & Wards Linkage Clustering in EG with the following variables:


3 demographic variable(Binary) - Gender(Female/Non-Female), Language(English/French), Ethnicity(Ethnicity1/Ethnicity2).

9 variables that reflect %Sales spent in each of the 9 departments (sum up to 100%.)

15 more Sales and visit pattern variables.


This is for a retail company and one area they particular interested is %spending by department.


But the segment I got seem to be mostly split along the demographic line, the 6 segment is as follow:

Seg 1:  100% French Speaking

Seg 2:  100% Ethnicity1 Female

Seg 3:  100% Ethnicity1 Male

Seg 4:  100% Ethnicity2 Female

Seg 5:  100% Ethnicity2 Male

Seg 6:   The rest


I suspect a huge reason this is  their department is gender-specific, like menswear, women wears, mens shoes, woman's shoe, cosmetics and etc.  And Stores in French speaking area have different assortment.


My question is:

1)  Should I remove the demographic variables during when I do the clustering?

2)  For dimension reduction, I used varclus instead of factor analysis, will this affect anything?

3)  I done Canonical Discriminant Analysis and got this.  How do I interrupt this?


Thanks in advance.

Ask a Question
Discussion stats
  • 0 replies
  • 1 in conversation