Professional Documents
Culture Documents
Report by:
Group - 13, Section B
Names:
Kasturi Vineel Chandra (2021SMT6668)
Noonsavath Sandeep (2021SMF6542)
Monika Mardi (2021SMF6516)
Kirish Jhamtani (2021SMF6609)
Introduction:
Ajanta Supermarket, our client, is a large-sized retail brand that has its offline stores operating
in the three major cities of Myanmar. It offers various products falling under the following
categories:
1. Electronic Accessories
2. Home and Lifestyle
3. Health and Beauty
4. Fashion Accessories
5. Food and Beverages
6. Sports and Travel
The client wants us to help them with the following broad business objectives:
Problem-Solving Process:
Problem Understanding -> Data Collection -> Data Understanding -> Data Cleaning ->
Exploratory Data Analysis -> Hypothesis Testing -> Regression Analysis -> Insights &
Recommendations
Consumer behavior
The sales are almost equal in all the product lines which states that the company is not stocking
any product which has less probability to get sold.
Customers bought 10 units at a time, which is most frequently seen when compared with the
other number of units bought in a single purchase. From this, it is evident that people
psychologically tend to buy 10 products at a time. The company can use this insight to attract
people who are buying 8 or 9 products at a time to make them buy 10 by offering a little
discount.
The average rating for food and beverages stands highest with 7.113, From the graph , it can be
inferred that all the accessories and products which falls under personal usage segment has
good rating except Home and life style which has 6.837 as average rating. The company can
perform better in home and lifestyle products which could help them to position themselves as a
brand which sells good products for personal usage.
Branch wise analysis
All their branches in three different cities sold same number of goods which means that all the
branches are equally performing which is a healthy sign for the company .
All the branches have sold almost equal number of products in different product lines except in
Nypitaw branch as there are least number of sales in the sports & travel, Home & lifestyle
products. Though sports & travel, Home and life style products in Naypyitaw branch accounts
for least number of sales, the average rating of those customers are above par.
The company has successfully turned almost 50% of its customers into members and company
can turn more into its members to make loyal customers.
Pearson’s Correlation Analysis:
Relationship between continuous variables, taken two at a time, can be determined by using
Correlation Analysis. The relationship can vary anywhere between Strongly Positive and
Strongly Negative.
In our case, the variables under consideration are - Unit Price, Quantity, Tax, Total Price,
Price Before Tax, Gross Income and Rating.
Subsequently, if we visualize the above results using a heatmap, the output will look something
like below:
Observations:
1. Rating has no relationship with every other continuous variable in the data
2. The strongest correlation exists between Quantity and Tax, Total Price, Price Before
Tax and Gross Income
3. The second highest correlation coefficient is between Unit Price and Tax, Total Price,
Price before Tax and Gross Income
Hypothesis Testing:
Hypothesis Statements:
Null Hypothesis (H0): Gender and Product Line are independent on each other
Alternate Hypothesis (H1): Gender and Product Line are dependent on each other
Null Hypothesis (H0): Customer Type and Product Line are independent on each other
Alternate Hypothesis (H1): Customer Type and Product Line are dependent on each other
Null Hypothesis (H0): City and Product Line are independent on each other
Alternate Hypothesis (H1): City and Product Line are dependent on each other
Null Hypothesis (H0): City and Payment Mode are independent on each other
Alternate Hypothesis (H1): City and Payment Mode are dependent on each other
Null Hypothesis (H0): Gender and Payment Mode are independent on each other
Alternate Hypothesis (H1): Gender and Payment Mode are dependent on each other
Chi-square Test:
Observations:
1. P-values in all the five cases are greater than 0.05 (significance level)
2. Calculated X-squared values in all the five cases are lesser than critical values
3. Thus, we cannot reject the null hypothesis in all the five cases
To find the least square regression line between Gross income (dependent) and unit price
(independent) separately for:
1. Electronic accessories
2. Food and beverages
3. Sports and travel
4. Home and lifestyle
Gross Estimated Slope Intercept Multiple R Significance
income- Unit Regression line R Square F
price
Generic Case
Introduction:
We have collected the transaction data of retail industry in Myanmar. We are particularly
interested in the Brick-and-Mortar stores. There are four companies operating in this space and
due to security reasons, we can’t disclose the company names. However, we have collected the
data related to customer spends, annual income, ratings, time of order etc. Our intent is to deep
dive into the data and find meaningful insights.
Normal 4.027102804
Member 3.98172043
Members might not be getting the value for their membership money, thus the rating is low.
Member 52.20430108
Normal 48.45794393
The industry has managed to bring a good chunk of customers into their membership program.
Moreover, we see that a member’s average spending score is higher compared to a normal
customer. This shows that we can expect revenue growth by converting normal customers into
members.
Here, we see that most customers shop between 5 to 9 PM and least customers shop between
7 to 9 AM. So, the staff can be employed in shifts, keeping the number of orders in mind.
Cash 4.038461538
Ewallet 3.934246575
Payments through Ewallet seems to have some trouble associated and that can be inferred
from the average rating associated with it. A better Ewallet transaction system could improve
the customer satisfaction significantly.
The graph shows that the spending score is high between 11 AM to 3 PM and is lowest
between 9 to 11 AM.
1 23 22 45
2 32 27 59
3 24 29 53
4 14 29 43
1 4.053333333
2 4.047457627
4 3.979069767
3 3.941509434
There is a significant gap observed between the ratings received by company 3 and 4 with
company 3.
One Way ANOVA Test
Null Hypothesis (H0): There is no difference in means of ratings between payment modes
Alternate Hypothesis (H1): There is difference in means of ratings between payment modes
Observations:
1. P-value is 0.411 (> 0.05), thus, the null hypothesis cannot be rejected
2. Thus, we do not have sufficient evidence to state that there is statistically significant
difference between payment modes
2. Between Spending Score and Payment Mode
Null Hypothesis (H0): There is no difference in means of spending scores between payment
modes
Alternate Hypothesis (H1): There is difference in means of spending scores between
payment modes
Observations:
1. P-value is 0.886 (> 0.05), thus, the null hypothesis cannot be rejected
2. Thus, we do not have sufficient evidence to state that there is statistically significant
difference between payment modes
Tools Used:
1. Microsoft Excel
2. RStudio
References:
1. Kaggle
2. Sthda
3. Towards Data Science
4. Statistics Globe
5. Guru99
6. Tutorials Point
Price
Customer Product Unit Quan Total before gross margin gross
Branch City type Gender line price tity Tax 5% Price Payment tax percentage income Rating
Health
and
A Yangon Member Female beauty 74.69 7 26.141 548.97 Ewallet 522.83 4.761904762 26.1415 9.1
Electroni
c
Naypyit accessor
C aw Normal Female ies 15.28 5 3.82 80.22 Cash 76.4 4.761904762 3.82 9.6
Home
and Credit
A Yangon Normal Male lifestyle 46.33 7 16.215 340.52 card 324.31 4.761904762 16.2155 7.4
Health
and
A Yangon Member Male beauty 58.22 8 23.288 489.04 Ewallet 465.76 4.761904762 23.288 8.4
Sports
and
A Yangon Normal Male travel 86.31 7 30.208 634.37 Ewallet 604.17 4.761904762 30.2085 5.3
Electroni
c
Naypyit accessor
C aw Normal Male ies 85.39 7 29.886 627.61 Ewallet 597.73 4.761904762 29.8865 4.1
Electroni
c
accessor
A Yangon Member Female ies 68.84 6 20.652 433.69 Ewallet 413.04 4.761904762 20.652 5.8
Transaction Gender Age Annual Spending Score Company Rating Custom Payment Time of
ID Income (k$) (1-100) er type Order