
✐
✐
“4137X˙CH05˙Akerkar” — 2007/9/8 — 11:01 — page 200 — #24
✐
✐
✐
✐
✐
✐
200 CHAPTER 5 Clustering
C1
C2
C3
C4
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Visits
Items
Variables
Spending
Normalized Values
Figure 5.20 Results of clustering customers from Gazelle.com
that the items and spending is an indication of spending potential and visits as a measure of
loyalty, we can label the clusters as follows:
C1: Semi-loyal, big spender
C2: Loyal, moderate spender
C3: Semi-loyal, moderate spender
C4: Infrequent, low spender
This analysis is based on limited data and analysis, and thus is not considered to be very mean-
ingful. However, it gives us some insight into how clustering can be useful ...