Performance Lawn Equipment (PLE) recently received data from a 3

rd
party survey of purchasing managers, their
customers, related to predicting the level of business and service. We will use Hierarchical Clustering a data
mining technique to reduce the data into groups and analyze the ratings for the distinct groups. We will also use
the data to determine and improve satisfaction drivers and usage level.
The survey performed has seven questions to evaluate the level of service from PLE to the purchasing managers.
Each manager was asked to rate PLE on a scale from 0 to 10 (poor to excellent) graphically to evaluate: delivery
speed, price level, price flexibility, manufacturing image, overall service, sales force image, and product quality.
With Hierarchical Clustering we will group purchasing managers into groups with similar responses. Our goal is
to be able to evaluate segments of our customer base to improve our customer relationships.
We will start with the data responses and using Hierarchical Clustering with average group linkage (using the
average of each group from all data points) to produce a dendrogram so we can properly evaluate the most
relevant number of group divisions. The dendrogram is a backwards (top to bottom) division of the data from
one large group to each individual observations showing the division of data into similar groups. For our analysis
we will determine a number of groups by drawing a horizontal line through the response level of 2.75, which
divides the data into 4 groups (see below):



Group 1 (which consists of)
1 11 9 28 12 14 5 7 13 18 15 23 17 24 21 2 19 6 25 8 16 4 20 26 22 29 3 10 27 30
0
0.5
1
1.5
2
2.5
3
3.5
4
D
i
s
t
a
n
c
e

Dendrogram(Average group linkage)
Row Id.

Cluster Id
Sub
Cluster Id
Delivery
speed
Price level
Price
flexibility
Manufacturing
image
Overall
service
Salesforce
image
Product
quality
1

1 1 4.1 0.6 6.9 4.7 2.4 2.3 5.2
5

1 5 6 0.9 9.6 7.8 3.4 4.6 4.5
7

1 7 4.6 2.4 9.5 6.6 3.5 4.5 7.6
9

1 9 5.5 1.6 9.4 4.7 3.5 3 7.6
11

1 11 2.4 1.6 8.8 4.8 2 2.8 5.8
12

1 12 3.9 2.2 9.1 4.6 3 2.5 8.3
14

1 7 3.7 1.5 8.6 5.7 2.7 3.7 6.7
15

1 14 4.7 1.3 9.9 6.7 3 2.6 6.8
19

1 7 5.3 1.4 9.7 6.1 3.3 3.9 6.8
20

1 14 4.7 1.3 9.9 6.7 3 2.6 6.8
28

1 7 5.2 1.3 9.7 6.1 3.2 3.9 6.7
38

1 12 4 0.9 9.1 5.4 2.4 2.6 7.3
42

1 5 5.9 0.9 9.6 7.8 3.4 4.6 4.5
49

1 9 5.8 0.2 8.8 4.5 3 2.4 6.7
51

1 1 3.7 0.7 8.2 6 2.1 2.5 5.2
58

1 28 5.4 2.5 9.6 5.5 4 3 7.7
63

1 12 4.1 1.1 9.3 5.5 2.5 2.7 7.4
66

1 12 3.7 1.4 9 4.5 2.6 2.3 6.8
67

1 7 4.2 2.5 9.2 6.2 3.3 3.9 7.3
74

1 9 5.2 1.3 9.1 4.5 3.3 2.7 7.3
76

1 12 4.2 2.4 9.4 4.9 3.2 2.7 8.5
77

1 1 3.8 0.8 8.3 6.1 2.2 2.6 5.3
85

1 11 2.6 3 8.5 6 2.8 2.8 6.8
87

1 11 2.4 2.9 8.4 5.9 2.7 2.7 6.7
90

1 7 4.3 2.5 9.3 6.3 3.4 4 7.4
95

1 1 4 0.5 6.7 4.5 2.2 2.1 5
97

1 9 6.1 0.5 9.2 4.8 3.3 2.8 7.1
100

1 11 2.5 1.8 9 5 2.2 3 6
(Row Id.) Customer identifier
has varying opinions about PLE, but more on the positive side. The delivery speed and price flexibility are
ranked high compared to other customers. They believe manufacturing and salesforce image is good. They are
not as satisfied as other customers with overall service and quality. They also feel the price is the biggest
obstacle in the business relationship.
Group 2 (which consists of)
Row Id.

Cluster Id
Sub
Cluster Id
Delivery
speed
Price level
Price
flexibility
Manufacturing
image
Overall
service
Salesforce
image
Product
quality
2

2 2 1.8 3 6.3 6.6 2.5 4 8.4
4

2 4 2.7 1 7.1 5.9 1.8 2.3 7.8
6

2 6 1.9 3.3 7.9 4.8 2.6 1.9 9.7
8

2 8 1.3 4.2 6.2 5.1 2.8 2.2 6.9
17

2 16 3.2 4.1 5.7 5.1 3.6 2.9 6.2
23

2 19 3 4 9.1 7.1 3.5 3.4 8.4
24

2 20 2.4 1.5 6.7 4.8 1.9 2.5 7.2
27

2 20 2.4 1.5 6.6 4.8 1.9 2.5 7.2
32

2 19 2.8 3.8 8.9 6.9 3.3 3.2 8.2
36

2 6 1.8 3.3 7.5 4.5 2.5 2.4 7.6
39

2 22 0 2.1 6.9 5.4 1.1 2.6 8.9
40

2 20 2.4 2 6.4 4.5 2.1 2.2 8.8
41

2 6 1.9 3.4 7.6 4.6 2.6 2.5 7.7
45

2 6 2 2.6 6.5 3.7 2.4 1.7 8.5
52

2 25 2.6 4.8 8.2 5 3.6 2.5 9
54

2 26 2.8 2.4 6.7 4.9 2.5 2.6 9.2
56

2 19 2.9 2.6 7.7 7 2.8 3.6 7.7
60

2 25 2.3 4.5 8 4.7 3.3 2.2 8.7
64

2 16 3 3.8 5.5 4.9 3.4 2.6 6
65

2 29 1.1 2 7.2 4.7 1.6 3.2 10
68

2 8 1.6 4.5 6.4 5.3 3 2.5 7.1
70

2 25 2.3 3.7 8.3 5.2 3 2.3 9.1
75

2 4 3 2 6.6 6.6 2.4 2.7 8.2
79

2 29 1 1.9 7.1 4.5 1.5 3.1 9.9
83

2 2 1.6 2.8 6.1 6.4 2.3 3.8 8.2
84

2 6 2.3 3.7 7.6 5 3 2.5 7.4
86

2 6 2.5 3.1 7 4.2 2.8 2.2 9
88

2 6 2.1 3.5 7.4 4.8 2.8 2.3 7.2
89

2 4 2.9 1.2 7.3 6.1 2 2.5 8
91

2 19 3 2.8 7.8 7.1 3 3.8 7.9
94

2 26 1.9 2.7 5 4.9 2.2 2.5 8.2
96

2 22 0.6 1.6 6.4 5 0.7 2.1 8.4
98

2 26 2 2.8 5.2 5 2.4 2.7 8.4
99

2 4 3.1 2.2 6.7 6.8 2.6 2.9 8.4
(Row Id.) Customer identifier
is the group that is most dissatisfied with PLEs partnership. They only rank PLE in the top half of customer
responses in product quality. Price, price flexibility, manufacturing image, and salesforce image are all ranked
below the average. They are most dissatisfied with delivery speed and overall service.
Group 3 (which consists of)
Row Id.

Cluster Id
Sub
Cluster Id
Delivery
speed
Price level
Price
flexibility
Manufacturing
image
Overall
service
Salesforce
image
Product
quality
3

3 3 3.4 5.2 5.7 6 4.3 2.7 8.2
10

3 10 4 3.5 6.5 6 3.7 3.2 8.7
30

3 10 4.1 3.7 5.9 5.5 3.9 3 8.4
31

3 10 3 3.2 6 5.3 3.1 3 8
34

3 10 3.4 3.7 6.4 5.7 3.5 3.4 8.4
37

3 10 3.6 4 5.8 5.8 3.7 2.5 9.3
48

3 10 3.4 3.9 5.6 5.6 3.6 2.3 9.1
53

3 10 4.5 4.1 6.3 5.9 4.3 3.4 8.8
57

3 27 4.9 4.4 7.4 6.9 4.6 4 9.6
71

3 3 3.6 5.4 5.9 6.2 4.5 2.9 8.4
82

3 30 3.4 4.6 5.5 8.2 4 4.4 6.3
93

3 30 3.1 4.2 5.1 7.8 3.6 4 5.9
(Row Id.) Customer identifier
is the group that seems well pleased with their business relationship with PLE. They rank at the top of the
customer surveys in price level, manufacturing image, overall service, salesforce image, and product quality.
Delivery speed and price flexibility are also ranked above average.
Group 4 (which consists of)
Row Id.

Cluster Id
Sub
Cluster Id
Delivery
speed
Price level
Price
flexibility
Manufacturing
image
Overall
service
Salesforce
image
Product
quality
13

4 13 2.8 1.4 8.1 3.8 2.1 1.4 6.6
16

4 15 3.4 2 9.7 4.7 2.7 1.7 4.8
18

4 17 4.9 1.8 7.7 4.3 3.4 1.5 5.9
21

4 13 3.3 0.9 8.6 4 2.1 1.8 6.3
22

4 18 3.4 0.4 8.3 2.5 1.2 1.7 5.2
25

4 21 5.1 1.4 8.7 4.8 3.3 2.6 3.8
26

4 21 4.6 2.1 7.9 5.8 3.4 2.8 4.7
29

4 15 3.5 2.8 9.9 3.5 3.1 1.7 5.4
33

4 21 5.2 2 9.3 5.9 3.7 2.4 4.6
35

4 13 2.4 1 7.7 3.4 1.7 1.1 6.2
43

4 17 4.9 2.3 9.3 4.5 3.6 1.3 6.2
44

4 21 5 1.3 8.6 4.7 3.1 2.5 3.7
46

4 17 5 2.5 9.4 4.6 3.7 1.4 6.3
47

4 23 3.1 1.9 10 4.5 2.6 3.2 3.8
50

4 24 5.4 2.1 8 3 3.8 1.4 5.2
55

4 18 3.8 0.8 8.7 2.9 1.6 2.1 5.6
59

4 21 4.3 1.8 7.6 5.4 3.1 2.5 4.4
61

4 23 3.1 1.9 9.9 4.5 2.6 3.1 3.8
62

4 21 5.1 1.9 9.2 5.8 3.6 2.3 4.5
69

4 24 5.3 1.7 8.5 3.7 3.5 1.9 4.8
72

4 24 5.6 2.2 8.2 3.1 4 1.6 5.3
73

4 15 3.6 2.2 9.9 4.8 2.9 1.9 4.9
78

4 15 3.3 2.6 9.7 3.3 2.9 1.5 5.2
80

4 17 4.5 1.6 8.7 4.6 3.1 2.1 6.8
81

4 24 5.5 1.8 8.7 3.8 3.6 2.1 4.9
92

4 17 4.8 1.7 7.6 4.2 3.3 1.4 5.8
(Row Id.) Customer identifier
has varying opinions about PLE, but mostly in the below average region. They only rank PLE above average in
price level and overall service. The purchasing managers feel PLE is slightly below average in delivery speed.
The rank price flexibility, manufacturing image, salesforce image, and product quality at the bottom of the
customer groups.
Given this data and analysis PLE knows the areas they need to focus on with each customer group to improve
the business relationship. It also shows how PLE is perceived in the industry by the select purchasing managers.
The group statistics show several different levels of perceptions regarding PLE. For instance, the group average
range in price (1.5 to 4.15) differs dramatically than the averages in quality ranging (5.18 to 8.25). This fact
illustrates PLE is a premium company with the focus of the company being quality. With this knowledge PLE will
be able to target the purchasing managers dissatisfaction and areas of concern while maintaining better
relationships with others in their core business mission.
We are also going to explore the usage and satisfaction level of the purchasing firms by evaluating different
characteristics from the survey. The characteristics are: size of the firm, purchasing structure (centralized or
decentralized), industry classification (resale or nonresale), and buying type (new, modified rebuy, or straight
rebuy). We will use correlation analysis to evaluate the driving factor in usage level and satisfaction level to
derive a cause-and-effect model

Correlation Analysis:
Usage Level
Satisfaction
Level Size of firm
Purchasing
Structure Industry
Buying
Type
Usage Level 1

Satisfaction Level 0.710697543 1

Size of firm -0.365171196 -0.456556445 1
Purchasing
Structure -0.402517394 -0.466362543 0.816496581 1

Industry -0.080503479 -0.074007154 9.06493E-18 0.04 1

Buying Type 0.828454966 0.709427536 -0.594088526 -0.5820855 0.048507125 1

As you can see from the above data, usage level is highly correlated with satisfaction level (as you expect) and
buying type. It is also strongly correlated in the negative direction with the size of the firm and the purchasing
structure. The satisfaction level is strongly correlated with buying type and negatively with the size of the firm
and the purchasing structure. The size of the firm and purchasing structure are very strongly correlated with
buying type also showing a relationship. It doesnt appear Industry type is correlated with any other factors
evaluated, so we did not add it to the Cause-and-Effect evaluation. Finally the purchasing structure is negatively
correlated with buying type. This would leave you to believe that buying type is a driving factor in both usage
level and satisfaction level. The data also show a significant negative relationship between the size of the firm,
purchasing structure with both usage level and satisfaction level. Our evaluation leads to the chart below:
Cause-and-Effect Chart from Correlations:










In conclusion the size of the firm determines purchasing structure which leads to buying type. These 3 factors
most effect the satisfaction level, which is the main determining factor in usage level.
PLE now has groups to further analyze and work to improve performance. They also have a Cause-and-Effect
flow chart determining Usage level, which is the determining factor in PLE success. With the information
provided PLE can implement a business plan to improve business relationships and increase the amount of
goods being sold to purchasing managers, which will lead to future business success.
Usage Level
Satisfaction level
Purchasing
structure
Buying type
Size of Firm