Scribd
Upload
27 views

BAM Analysis

The document summarizes various analyses performed on survey data from 60 respondents about their preferences for apparel products. Hierarchical cluster analysis identified 2 main clusters of respondents. K-means cluster analysis assigned most respondents to Cluster 2. Multidimensional scaling showed comfort, brand and quality were most important factors over color, trends and price. Conjoint analysis was used to understand preferences for different attribute combinations of apparel products.

Uploaded by

pratik panchal
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
27 views

BAM Analysis

The document summarizes various analyses performed on survey data from 60 respondents about their preferences for apparel products. Hierarchical cluster analysis identified 2 main clusters of respondents. K-means cluster analysis assigned most respondents to Cluster 2. Multidimensional scaling showed comfort, brand and quality were most important factors over color, trends and price. Conjoint analysis was used to understand preferences for different attribute combinations of apparel products.

Uploaded by

pratik panchal
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 15

Course Title – Business Analytics for Marketing

Course Instructor – Dr. HariBabu Singh

Type of Assignment – Group Assignment

Student Names and PRN NO –


Vedangi Sawant – 21021241133
Nandita Rai – 21021241089
Pratik Panchal – 21021241105

DECLARATION: -
We declare that this continuous evaluation component is our original work.
We have not copied from any source, or any other students work. We
have acknowledged the source, if any content is used in completion of this
continuous evaluation component. We are responsible for any
discrepancies/unfair means content/copied material presented in the
assignment/case analysis.

CLUSTER ANALYSIS

Cluster analysis is a class of techniques that is used to classify objects or


cases into relative groups called clusters. It is also called numerical
taxonomy or classification analysis. It is a multivariate approach for
identifying objects or individuals that are similar to one another in some
aspect.
Cluster Analysis approach has been used in marketing for various
purposes. Segmentation of consumers in cluster analysis is used on the
basis of benefits sought from the purchase of the product.
Objects or individuals are assigned to groups so that there is great
similarity within groups and much less similarity between groups. The
cluster should have high internal which is within- cluster homogeneity and
external which is between-cluster heterogeneity. It can be used to identify
homogeneous groups of buyers.

Cluster analysis involves –


1. Formulating a problem,
2. Selecting a distance measure,
3. Selecting a clustering procedure,
4. Deciding the number of clusters,
5. Interpreting the profile clusters and finally,
6. Assessing the validity of clustering.

Objective – To understand the usage and preference of Indian consumers


towards ‘Apparels’ category.

PRODUCT – Apparels
In this case, we used –
1. Hierarchical method
2. Non-hierarchical method – K means clustering

We took a total of 60 respondents for analysis and the following steps were
followed for the analysis –

HEIRARCHIAL ANALYSIS
We first obtain responses from the people via survey on apparels and then
import the data to the SPSS Software.

Then we run the Hierarchical Analysis to obtain the results as follows –


1. Agglomeration Schedule – Like using Ward’s Linkage method, we got
Agglomeration schedule which provides us what new information is
provided by each additional cluster.
This is basically numerical summary of the clusters. As we can see there
are stages and clusters. In stage 1, cases 36 and 37 are combined as
they have the smallest distance between them.
The coefficient column reports similarity and therefore the coefficient
values keep decreasing after each stage.

2. Case
3. Dendrogram – This shows the hierarchical relationship between the
objects. It is most commonly created as an output from hierarchical
clustering method. The main use of a dendrogram is to work out the
best way to allocate objects to clusters.
K MEANS CLUSTER ANALYSIS
We get the following results after K means cluster analysis –

1. The initial cluster center table shows


the initial centers of each cluster.

2. The iteration history shows the


progress of the clustering process at
each step
3. The cluster membership table in K means
clustering shows us which case lies in
which clusters.

4. The Final cluster centers table shows us the characteristics of the


typical case for each cluster. These are computed as the mean for each
variable within each final cluster.
5. The Distances between Final cluster centers table reflects the Euclidean
distances between the final cluster centers. Euclidean distance is the
default distance measure. The greater the distances between the
clusters, the more dissimilarities are there. This table shows significant
differences between them.

6. The next analysis would be of the ANOVA table which tells us which
variables in our study contribute the most to our clustering solution.
The greater the F value, the greater the separation between clusters.

7. The number of Cases in each cluster is represented by the table given


below. As we can see from the table, cluster 2 has been assigned the
largest number of cases as compared to cluster 1 out of the total cases
of 60.
Conclusion –

Therefore, from the final results of clustering we can conclude that Area as
a factor is not contributing in the clustering of the respondents. Other
factors which are not significantly contributing in clustering of respondents
are –

The factors present in the ANOVA table are the only ones who significantly
contribute to the clustering.
As we can see that the respondents of the apparel survey are finally
divided into 2 clusters where cluster no. 2 being with the highest
respondents. The clustering was done on the basis of respondent’s
similarities and dissimilarities.
The Apparel brands should focus more on what type of respondents belong
to each cluster while focusing more on the 1st cluster as it has the
maximum respondents with similarities. This way they will cater to a
larger customer segment and can come up with new types of marketing
strategies to target these particular group of customers.

MULTIDIMENSIONAL SCALING (MDS)

Multidimensional Scaling (MDS) is a series of process which represents the


respondent’s perception and preferences. The relationship among various
factors is represented geometrically in multidimensional space. These
spaces are known as spacial maps. MDS is used in detecting meaningful
dimensions which allows the researchers to explain the similarities and
dissimilarities between the objects.

MDS helps is assessing the positions of a product on the perceptual map with
other competitors.

The points to be considered while performing MDS analysis are –


1. Selection of the objects to be evaluated
2. Similarity on the basis of overall similarity or dissimilarity
3. Aggregate evaluation of the objects

To know how many pairs will be formed, we use the formula n(n-1)/2,
where n is the number of objects.

Stress and squared correlation (RSQ) in distances

RSQ values are the proportion of variance of the scaled data


(disparities)
in the partition (row, matrix, or entire data) which
is accounted for by their corresponding distances.
Stress values are Kruskal's stress formula 1.

For matrix
Stress = .01211 RSQ = .99867

Configuration derived in 2 dimensions

Stimulus Coordinates

Dimension

Stimulus Stimulus 1 2
Number Name

1 Price 1.2777 1.0974


2 Trends -1.7591 .7880
3 Comfort -.6295 .4846
4 Color -.6340 -1.5211
5 Quality 1.1511 -.6017
6 Brand .5939 -.2473

Optimally scaled data (disparities) for subject 1

1 2 3 4 5 6

1 .000
2 3.053 .000
3 2.026 1.170 .000
4 3.234 2.569 2.026 .000
5 1.737 3.234 2.026 2.026 .000
6 1.468 2.571 1.468 1.737 .660 .00

Conclusion:
By the above results and and plots we can say that Comfort, Brand, and
Quality are towards the origin compared to colour, trends and price factor.
Hence customers give more preference to comfort, brand, and Quality of
apparels.
The R square value is 0.998 which is greater than 80% hence the model is fit.
The Stress value is 0.1207 which is less than 0.2 and hence within limits.

CONJOINT ANALYSIS

Conjoint Analysis is a statistical tool or a method of product and pricing


research which uncovers customers preferences and uses that information
to select product features, assess sensitivity to price, forecast market
shares and predict adoption of new products or services. Conjoint analysis
works by breaking a product or service down into its components (referred
to as attributes and levels) and then testing different combinations of
these components to identify consumer preferences.
Going further than simply asking respondents what they like in a product,
or what features they find most important, conjoint analysis employs a
more realistic approach: asking each respondent to choose between
potential product concepts (or alternatives) formed through the
combination of attributes and levels. These combinations are carefully
assembled into choice sets (or questions).

We have done the analysis on each respondents, so there are total 60


subjects out of which is showcased below.

You might also like