FAST SCHOOL OF MANAGEMENT

DATA ANALYSIS FOR BUSINESS -II PROJECT (FALL 2024)

Consumer Insights in the Supermarket Industry: Analyzing Sales

Trends, Customer Behavior, and Product Preference

Submitted To:

Ms. Aroosa Safdar

Section:

BAF-3A

Submitted By:

M. Hussnain Raza (23I-4528)

Abdul Moeez (23L-4565)

Bilal khan(23L-4547)

Sarim Zakaria (23L-4559)

M Ibrahim khan (23L-4517)

 ABSTRACT

This report presents an in-depth analysis of supermarket sales data, focusing on key business
metrics such as Unit Price, Total Sales, Gross Income, and Rating. The main objective is to
explore the relationships between these metrics and various customer characteristics, such as
gender and customer type. The study answers nine key research questions through statistical
techniques including normality testing, correlation analysis, and hypothesis testing (t-tests,
ANOVA, chi-square tests).

The analysis investigates whether the data for these variables is normally distributed, whether
there are gender-based differences in ratings, and whether sales performance varies across
branches. It also explores correlations between Rating and Gross Income, and between Unit
Price and Quantity, as well as testing the independence of Product Lines from gender and the
relationship between Customer Type and Payment Method.

The findings reveal that all tested variables fail normality tests, indicating non-normal
distributions. There is no significant difference in ratings between males and females, and
total sales do not vary significantly across branches. The correlation between Rating and
Gross Income is found to be insignificant, as is the correlation between Unit Price and
Quantity. Additionally, no significant association exists between Product Line and Gender,
and Customer Type does not significantly influence Payment Method. The mean Rating is
significantly different from 8, and the mean Total Sales is greater than 300.
 ACKNOWLEDGEMENTS

We would like to express our sincere gratitude to Maam Aroosa Safdar, our instructor, for
Her constant support, guidance, and valuable feedback throughout this project. Her expertise
in statistical methods and attention to detail were critical in refining our analysis and ensuring
that all methodologies were correctly applied.

We would also like to thank Maam Sadia Manzoor Malik for her invaluable mentorship
and thoughtful suggestions, which greatly enhanced the quality of our report. Her guidance
helped us deepen our understanding of data analysis techniques and apply them effectively to
address the research questions.

Additionally, we are incredibly grateful to our group members for their collaboration and
hard work throughout the project. Their dedication, ideas, and teamwork made this project
possible, and we truly appreciate the support and effort each member contributed.

We are grateful for the encouragement, patience, and dedication of our instructors and group
members, which made this project a rewarding learning experience. Without their assistance,
this report would not have been possible.
 LIST OF TABLES

Table 1: Branch Frequency Table 13

Table 2:Customer type frequency Table 13
Table 3:Gender frequency Table 14
Table 4:Product line Frequency Table 14
Table 5: Payment Frequency table 15
Table 6: Descriptive Stats (Unit Price, Quantity, Total, Gross Income, Rating 15
Table 7:Descriptive of unit price, gross income ,rating 18
Table 8:Descrpitive of tax, quantity and COGS 20
Table 9:Cross tabulation of Gender and Product line 21
Table 10:Crosstabulation of Customer type and payment 22
Table 11:Normality Test (Unit Price, Total, Gross Income, Rating 27
Table 12:Normality Test (Tax, Quantity, COGS) 27
Table 13:T-Test (Gender & Rating) 29
Table 14:ANOVA (Total Sales Across Branches) 30
Table 15:Pearson Correlation (Rating & Gross Income) 32
Table 16:Pearson Correlation (Unit Price & Total) 33
Table 17:Chi-Square (Customer Type & Payment Method) 35
Table 18:Regression Model between unit price and total 39
 LIST OF FIGURES

Figure 1: Pie Chart of Branch 18

Figure 2: Bar chart product line by Gender 19
Figure 3: Histogram of Rating 20
Figure 4: Histogram of Quantity 21
Figure 5: Histogram of tax5 21
Figure 6: Scatter with Fit line of Rating by grossincome 28
Figure 7: Simple Scatter with fit line of Total by unit price 30
 LIST OF ABBREVIATIONS

Df: Degrees of freedom

H0: Null Hypothesis

H1: Alternative Hypothesis

IBM: International Business Machines Corporation

IQR: Interquartile Range

N: Sample size

SPSS: Statistical Package for the Social Sciences

1 INTRODUCTION
The supermarket industry is an essential part of modern economies, serving as a central hub
for consumer goods. With increasing competition and shifting consumer preferences,
understanding the factors that drive sales, customer satisfaction, and branch performance has
become crucial for supermarket chains. Retailers are increasingly turning to data analytics to
make informed decisions about marketing strategies, inventory management, and customer
engagement. By leveraging transactional and demographic data, businesses can uncover
insights into purchasing behavior, operational efficiency, and revenue generation.

This study focuses on a detailed analysis of supermarket sales data to answer critical
questions about customer satisfaction, sales performance across branches, and the
relationships between various operational metrics. The research utilizes SPSS to conduct
descriptive and inferential statistical analyses on key variables, including ratings, sales totals,
gross income, and unit prices. By examining these factors, this study provides actionable
insights for retail management, helping to enhance decision-making processes.

The overarching aim of the research is to evaluate both individual and combined effects of
various factors on sales performance and customer satisfaction. For instance, do male and
female customers rate their experiences differently? How does branch location influence total
sales? Can pricing strategies predict gross income? These are just a few of the pertinent
questions that this study seeks to address.

Literature Review

Customer satisfaction has long been recognized as a key driver of business success, with
satisfied customers more likely to become repeat buyers and advocates for the brand (Kotler
& Keller, 2016). In the supermarket context, satisfaction ratings provide valuable insights
into consumer perceptions of service quality, product availability, and pricing, all of which
directly impact customer loyalty. Gender differences in consumer behavior have been well-
documented, with studies indicating that women tend to focus more on the overall shopping
experience, including customer service and convenience, while men are often more goal-
oriented, seeking efficiency and specific products (Babin et al., 2005). This study investigates
whether such behavioral differences influence satisfaction ratings in the supermarket setting.
Furthermore, branch-level performance is influenced by various factors such as foot traffic,
local competition, and socio-economic demographics, which can lead to regional disparities
in sales performance (Fisher et al., 2017). The analysis of branch-specific sales is essential
for identifying performance trends and operational improvements. In terms of pricing,
research suggests that a well-calibrated pricing strategy can optimize revenue without
alienating price-sensitive customers, and understanding the relationship between unit price
and total sales is vital for supermarket profitability (Nagle et al., 2016). Additionally, the rise
of digital and contactless payment systems has shifted consumer preferences, with younger
customers more likely to choose digital payment methods (Choi et al., 2020). This study
explores how payment preferences vary by customer type (e.g., members versus walk-ins).
Finally, hypothesis testing is an essential tool in retail analytics, enabling businesses to test
assumptions and make data-driven decisions. By using hypothesis testing, this study aims to
draw insights regarding differences in customer satisfaction, sales performance across
branches, and correlations between operational metrics such as gross income and customer
ratings.
1.1 OBJECTIVES OF THE STUDY
The main objective of this study is to analyze supermarket sales data to gain insights into
customer behavior, sales performance, and operational efficiency. The study focuses on
several specific objectives.

First, it aims to assess the normality of key variables, including unit price, total sales, gross
income, and customer ratings. This helps determine whether the data follows a normal
distribution, which is essential for selecting appropriate statistical tests.

Second, the study seeks to investigate if there are significant gender-based differences in
customer ratings. This will help understand whether male and female shoppers have distinct
satisfaction levels, which could influence store policies and service improvements.

Third, the analysis aims to compare sales performance across branches, identifying any
significant variations in sales figures and uncovering the factors influencing branch-level
performance, such as location or demographics.

Another objective is to explore the relationship between customer ratings and gross income.
This will determine if higher customer satisfaction correlates with increased revenue, helping
to assess whether improving customer experience impacts financial performance.

The study will also examine the correlation between unit price and total sales. Understanding
this relationship can guide pricing strategies to optimize sales and revenue.

Additionally, the study will test whether the distribution of product lines is independent of
gender, which could help in customizing product offerings based on customer demographics.

The analysis will also explore the association between customer type (e.g., member vs. walk-
in) and payment method. This will help determine if customer preferences for payment
methods vary based on their type, which can guide payment system optimization.

Lastly, the study will test hypotheses regarding the mean rating and the mean of total sales.
This will provide insights into customer satisfaction levels and the overall sales performance
of the supermarket.

By achieving these objectives, the study aims to generate actionable insights that can improve
supermarket operations, enhance customer satisfaction, and drive business success.
2 DATA COLLECTION & METHODOLOGY
The data used in this study was obtained from an online dataset consisting of supermarket
sales transactions. This dataset, provided in CSV format, includes detailed information on
various aspects of supermarket sales such as unit price, total sales, gross income, customer
ratings, gender, branch location, payment method, and customer type. These variables are
essential for addressing the research questions and objectives of the study. The data was
sourced from an open online repository designed for research purposes, offering valuable
insights into customer behaviour, sales performance, and operational factors affecting
supermarket sales.

To analyse this data, SPSS (Statistical Package for the Social Sciences) software was used,
which is a widely recognized tool in business and social science research. The initial step in
the methodology was data preparation and cleaning. The dataset was imported into SPSS and
reviewed for missing values, which were either imputed or excluded depending on their
nature. Additionally, some categorical data was transformed into a numerical format for
analysis, such as converting binary variables like "yes/no" responses into corresponding
numerical values.

Descriptive statistics were then applied to summarize the key features of the dataset.
Frequencies and percentages were calculated for categorical variables such as gender,
payment method, and customer type. For continuous variables like unit price, total sales, and
customer ratings, measures of central tendency (mean, median, mode) and dispersion
(standard deviation, range) were computed. To examine the relationship between categorical
variables, cross-tabulation was used, such as analyzing gender against product line. Various
visualizations, including bar charts, pie charts, and histograms, were created to illustrate the
distribution of key variables.

Next, normality testing was conducted for continuous variables to determine if the data
followed a normal distribution. This was done using the Shapiro-Wilk test, along with visual
methods such as histograms and Q-Q plots. Normality testing is a crucial step in deciding
which statistical tests are appropriate for hypothesis testing.

For hypothesis testing, several statistical techniques were employed. T-tests and ANOVA
were used to test for differences in means, such as comparing customer ratings between
genders and analyzing total sales across different branches. Chi-square tests for independence
were conducted to explore the relationships between categorical variables, such as whether
there was an association between customer type and payment method, or if the distribution of
product lines was independent of gender. Additionally, correlation analysis was used to
investigate relationships between continuous variables like unit price and total sales.

Throughout the analysis, SPSS software was utilized for all statistical procedures, including
data manipulation, hypothesis testing, and the creation of graphical representations of the
results.
3 ANALYSIS & RESULTS
3.1 Descriptive statistics
Frequency tables:

Branch

Interpretation: The transactions in the dataset are fairly evenly distributed across the three
branches, with Branch A accounting for 34.0% (340 out of 1000) of the transactions, Branch B with
33.2% (332 out of 1000), and Branch C with 32.8% (328 out of 1000). This suggests that customer
activity is relatively balanced across the branches, with no single branch dominating the sales.

● Customer type

Interpretation: The customer base is almost evenly split between members and non-members, with
50.1% (501 out of 1000) being members and 49.9% (499 out of 1000) being non-members. This
suggests that both loyal customers and occasional buyers represent nearly equal proportions of the
total customer pool.

● Gender

Interpretation: The gender distribution of customers is almost perfectly equal, with 50.1% (501 out
of 1000) female and 49.9% (499 out of 1000) male customers. This balanced gender representation
means that any analysis on gender-based preferences or satisfaction can be generalized, without any
gender being overrepresented.

● Product line
Table 4 Product line Frequency Table

Interpretation: Product preferences in the supermarket appear to be well-distributed, with some

categories showing higher sales than others. The most popular product line is Fashion Accessories,
accounting for 17.8% of total transactions (178 out of 1000), closely followed by Food and Beverages
at 17.4% (174 out of 1000), and Electronic Accessories at 17.0% (170 out of 1000). On the other
hand, Health and Beauty had the lowest representation at 15.2% (152 out of 1000). These percentages
suggest that there is a diverse customer interest in various product lines, but fashion-related items and
food remain dominant.

● Payment

Interpretation: In terms of payment methods, the E-wallet is the most widely used,
accounting for 34.5% of transactions (345 out of 1000), followed closely by Cash at 34.4%
(344 out of 1000) and Credit Card at 31.1% (311 out of 1000). The relatively high usage of
E-wallets suggests that the supermarket is catering to a tech-savvy customer base that prefers
digital payments

1. Descriptive table

● Overall Descriptive Statistics

The average unit price in the supermarket is $55.67, with a standard deviation of
$26.49, indicating a significant variation in the prices of products. The minimum unit
price is $10.08, and the maximum is $99.96, reflecting the diverse range of products,
from low-cost items to premium ones.

● Quantity
 Customers purchase an average of 5.51 items per transaction, with a minimum of 1
item and a maximum of 10 items. This suggests that the typical customer makes a
moderate purchase, though there is room for both very small and relatively larger
transactions.

● Total

The average total for a transaction is $322.97, with a wide range (from $10.68 to
$1042.65), reflecting both smaller and larger basket sizes. The large standard
deviation of $245.89 suggests that there is considerable variation in transaction sizes,
with some customers purchasing only a few items while others make bulk purchases.

● Gross Income

The gross income per transaction averages at $15.38, with a standard deviation of
$11.71. The maximum is $49.65 and minimum is $0.51, and. The variance in gross
income indicates that while most transactions yield modest profit, some are more
profitable than others.

● Rating

The average customer satisfaction rating is 6.97, with a standard deviation of 1.72,
suggesting a generally positive but somewhat variable level of customer satisfaction.
The ratings range from 4 to 10, indicating that customers have a broad spectrum of
satisfaction, with the average tending towards the higher end of the scale

● unit price,

The mean is $55.67, ranging from $10.08 to $99.96, with a nearly symmetric distribution
(skewness of 0.007) and a flatter-than-normal peak (kurtosis of -1.219).

● Gross income

mean of $15.38, with a range from $0.51 to $49.65, reflecting variability and a right-skewed
distribution.

● Rating
The average rating is 6.973, with a standard deviation of 1.7186, indicating moderate
variation. The ratings are symmetrically distributed, as shown by the near-zero skewness
(0.009), while the negative kurtosis (-1.152) suggests a flatter distribution than normal

● Tax

Tax data, likely representing 5% of the total, mirrors the gross income distribution, with
comparable statistics. Overall, the data reveals symmetry in non-financial metrics like rating
and quantity, while financial metrics exhibit more variability and skewness, indicating
diverse transaction sizes.

● Quantity

The mean of items per transaction is 5.51, with values ranging from 1 to 10. This variable
also exhibits a symmetric distribution (skewness of 0.013) and a flat peak (kurtosis of -
1.216).

● COGS

cost of goods sold (COGS) averages $307.59, with a substantial variation (standard deviation
of 234.18) and a right-skewed distribution (skewness of 0.893), indicating that higher values
are less common.

2. Crosstabulations

● Gender vs. Product Line Preferences

Interpretation:

The gender distribution across product lines shows distinct preferences. Female customers are
more likely to purchase Fashion Accessories (96 transactions) and Food and Beverages (90
transactions), while male customers show stronger interest in Health and Beauty (88
transactions) and Electronic Accessories (86 transactions). This suggests that product
preferences are influenced by gender, with females gravitating towards fashion and food-
related items, while males lean more towards electronics and personal care products. This
insight could guide future marketing campaigns, product placement, and inventory
management to better serve gender-specific interests.

● Customer Type vs. Payment Method

Customer type * Payment Crosstabulation

Interpretation:

There is an interesting pattern in payment preferences between customer types. Members,

who make up 50.1% of the total transactions, are more likely to use Credit Card (172
transactions) as a payment method, while non-members, making up 49.9% of the total
transactions, prefer E-wallet (184 transactions). This could reflect the convenience and
loyalty incentives associated with credit card payments for members, while non-members
may prefer the flexibility and digital convenience offered by E-wallets. This finding suggests
a potential area for further investigation into customer loyalty programs and their influence
on payment preferences.

3. Charts

● Pie chart

Figure 1 Pie Chart of Branch

Interpretation;

The three branches' percentage distribution of counts—A, B, and C—is depicted in the pie
chart. The highest portion is held by Branch A, which is depicted in blue, at 34.00%,
followed closely by Branch B, which is shown in red, at 33.20%. At 32.80%, Branch C,
which is shown in green, has the smallest part. The little percentage differences show that the
counts are distributed rather evenly among the three branches, with none of them
substantially outweighing the others.

● Bar chart

Figure 2 Bar chart product line by Gender

Interpretation:
the chart presents a clustered bar count of product line purchases by gender, highlighting the
distribution of purchases across different categories for male and female customers. Fashion
Accessories and Health and Beauty show a clear dominance of female customers, suggesting
that these product lines are more popular among women. Conversely, Food and Beverages
and Sports and Travel have a higher count among male customers, indicating a stronger
preference for these categories among men. Electronic Accessories and Home and Lifestyle
show more balanced participation from both genders, with males slightly leading in
electronics and females slightly leading in home products. These trends reflect gender-based
preferences, which can be valuable for tailoring marketing strategies and product offerings to
each group.

● Histogram

⮚ Rating
Figure 3 Histogram of Rating

Interpretation:
The histogram displays the distribution of ratings, ranging from 4.0 to 10.0, with frequency
represented on the y-axis. The average rating is 6.973, and the standard deviation is 1.7186,
indicating a moderate spread of ratings around the mean. The distribution appears relatively
even, with most ratings falling between 5.0 and 9.0, which are the most commonly observed
scores. Peaks in frequency occur near specific values, such as 6.0 and 9.0, with
approximately 50 occurrences. Ratings at the lower and upper extremes, closer to 4.0 and
10.0, are less frequent. With a total of 1,000 ratings in the dataset, the histogram highlights a
balanced yet diverse pattern of ratings, cantered around the average.

⮚ Quantity
Figure 4 Histogram of Quantity

Interpretation: Quantity, shows an approximately uniform distribution. This means the

values are spread relatively evenly across the intervals, with no strong peaks or troughs. The
mean value is 6.97, indicating that the average quantity lies around this value. The standard
deviation is 1.718, which suggests moderate variability around the mean..

⮚ Tax

Figure 5 Histogram of tax5

Interpretation: Tax histogram, demonstrates a right-skewed distribution. The majority of tax

values are concentrated at lower amounts, with a few higher values creating a long tail. The
mean tax amount is 15.3794, but the skewed nature of the data indicates that most values are
below this average, with a few larger values pulling the mean upward. The standard deviation
of 11.7088 reveals significant variability in tax amounts.
3.2 Inferential statistics
Q1: Normality Testing of Unit Price, Total, Gross Income, and Rating, Tax, Quantity
and COGS

● Objective:
To determine whether the data for Unit Price, Total, Gross Income, and Rating, tax,
quantity and cogs follows a normal distribution. This is essential for deciding the
appropriate statistical tests to apply (parametric or non-parametric).

● Problem:
Many statistical techniques assume normality in the data. If these variables deviate
significantly from normality, it could lead to biased or incorrect conclusions in
subsequent analyses.

● HYPOTHESES TESTING:
Null Hypothesis: Data is Normal

Alternative Hypothesis: Data is not Normal

Level of Significance: α=0.05

Test Statistic: As sample size is greater than 50, we will use Kolmogorov test value.

Here our p value is 0.000.

P-value Comparison
Reject H 0 if p ≤ α
Here, p-value = 0.000 <α
Decision: Accept H1

Conclusion: Data is not Normal

Q2: Does "Rating" Differ Between Males and Females?

● Objective:
To assess if there is a significant difference in customer satisfaction ratings (Rating)
between male and female customers. This can help in understanding gender-based
variations in customer experiences.

● Problem:
Gender-based biases or differences in satisfaction may indicate areas where service
improvements are needed to cater to diverse customer groups effectively.

● HYPOTHESES TESTING:

Table 13T-Test (Gender & Rating)

Null Hypothesis: There is no significant difference in "Rating" between males and females

Alternative Hypothesis: There is a significant difference in "Rating" between males and

females

Level of Significance: α=0.05

Test Statistic:
T value:0.152
P value:.879
P-value Comparison
Reject H 0 if p ≤ α
Here, p-value = .879 > α
Decision: Accept H0

Conclusion: There is no significant difference in "Rating" between males and females

Q3: Does "Total Sales" Differ Across Branches?

● Objective:
To analyse whether the Total Sales varies significantly across the three supermarket
branches (A, B, C). This can provide insights into the performance and customer
behaviour at each location.

● Problem:
Disparities in sales performance across branches might reflect location-specific issues
such as marketing effectiveness, customer base, or branch-specific operational
challenges.

● HYPOTHESIS TESTING

Null Hypothesis: Total sales are the same across branches

Alternative Hypothesis: Total sales are not the same across branches.

Level of Significance: α=0.05

Test Statistic:

P value:0.413
P-value Comparison

Reject H 0 if p ≤ α

Here, p-value 0.413 > α

Decision: Accept H0

Conclusion: Total sales are the same across branches

Q4: Is There a Relationship Between "Rating" and "Gross Income"?

● Objective:
To investigate the association between customer satisfaction ratings (Rating) and the
profit earned (Gross Income). A positive relationship may suggest that satisfied
customers contribute to higher profits.

● Problem:
Understanding the link between satisfaction and profitability is critical for aligning
customer service strategies with financial outcomes. A weak or non-existent
relationship could challenge this assumption.

● HYPOTHESIS TESTING

Alternative Hypothesis: There is a significant correlation between "Rating" and "Gross

Income

Null Hypothesis: There is no significant correlation between "Rating" and "Gross Income

Level of Significance: 0.01

Test Statistic:

P value:0.250

P-value Comparison

Reject H 0 if p ≤ α
 Here, p-value .250 > α

Decision: Accept H0

The Pearson correlation coefficient (r) = -0.36, indicating a weak negative correlation.

Conclusion: There is no significant correlation between "Rating" and "Gross Income”.

Figure 6 Scatter with Fit line of Rating by grossincome

Q5: Is There a Significant Correlation Between "Unit Price" and "Total"?

● Objective:
To examine if there is a significant correlation between the price per unit (Unit Price)
and the total cost of purchases (Total). This can reveal how pricing strategies impact
overall sales value.

● Problem:
Mismanagement in pricing could result in diminished sales or overpricing, potentially
affecting customer loyalty and revenue.

● HYPOTHESIS TESTING

Correlations
Alternative Hypothesis: There is a significant correlation between "Rating" and "Gross
Income

Null Hypothesis: There is no significant correlation between "Rating" and "Gross Income

Level of Significance: 0.05

Test Statistic:

P value:0.000

P-value Comparison

Reject H 0 if p ≤ α

Here, p-value .0.000<α

Decision: Accept H1,

The Pearson correlation coefficient (r) = 0.634, indicating a moderate positive correlation.

Conclusion: There is a significant correlation between "unit price" and "Total’’

Figure 7 Simple Scatter with fit line of Total by unit price

Q6: Is There a Significant Association Between Customer Type and Payment?

 ● Objective:
To analyse the relationship between customer type (Member vs. Non-member) and
preferred payment methods (Cash, Credit Card, etc.). This insight is valuable for
enhancing the customer experience.

● Problem:
Understanding payment behaviour across customer types is essential for designing
loyalty programs and improving payment systems.

● HYPOTHESIS TESTING

Null Hypothesis: There is no significant association between customer type and payment

Alternative Hypothesis: There is significant association between customer type and

payment

Level of Significance: 0.05

Test Statistic:

P value:0.074

P-value Comparison

Reject H 0 if p ≤ α

Here, p-value .074 > α

Decision: Accept H0

Conclusion: There is no significant association between customer type and payment

Q7: Is the “Rating” Mean Equal to 8?

● Objective:
To test whether the mean customer satisfaction rating is statistically equal to 8. This
serves as a benchmark for evaluating overall customer satisfaction.
 ● Problem:
If the actual mean significantly deviates from 8, it may indicate a need for
improvements in service quality or customer experience.

● HYPOTHEIS TESTING

Null Hypothesis: Average is equal to 8

Alternative Hypothesis: Average is not equal to 8

Level of Significance: 0.05

Test Statistic:

P Value:0.0000
T Value: -18.903

P-value Comparison

Reject H 0 if p ≤ α

Here, p-value 0.000< α

Decision: Accept H1

Conclusion: Mean is not equal to 8

Q8: Is the Mean of Total Sales Greater Than 300?

● Objective:
To test whether the average total sales per transaction exceed 300. This can serve as a
key performance indicator for evaluating the supermarket’s revenue generation per
transaction.
 ● Problem:
Underperforming sales figures could signal inefficiencies in pricing, product
selection, or customer retention strategies.

● HYPOTHESIS TESTING

Null Hypothesis: The average of Total Sales is equal to 300.

Alternative Hypothesis: The average of Total Sales is greater than 300

Level of Significance: 0.05

Test Statistic:

T value:2.945
P Value: 0.03
P-value Comparison

Reject H 0 if p ≤ α

Here, p-value 0.03< α

Decision: Accept H1

Conclusion: The mean of Total Sales is greater than 300

REGRESSION:

Q9: Compute a regression model from the data, which predict Total from the Unit
Price.

Dependent Variable: Total

Independent Variable: unit price

From the table we can compute the following regression line,

Y= -4.582 + 5.884X

Total = -4.582+ 5.882(unit price).

From the table we can predict the following values as well,

Independent Variable Dependent Variable

7.5 39.553

12 66.002

4 Conclusion
Key indicators such unit price, total sales, gross income, and customer evaluations are
analysed in this study to provide insightful information about supermarket operations. We
looked at how these data related to things like branch performance, consumer type, and
gender using a thorough statistical [Link] findings indicate a balanced customer base
with no significant differences in satisfaction ratings or product preferences based on gender,
suggesting the supermarket’s offerings are well-suited to both male and female customers.
However, subtle trends, such as females favoring fashion accessories and males showing a
preference for electronics and personal care products, could be leveraged for more targeted
marketing strategies.
Sales performance across branches was consistent, with no significant variations detected,
reflecting effective standardized operational strategies. However, this uniformity might signal
missed opportunities to tailor branch-specific approaches to regional demands. Pricing
emerged as a critical factor, with a significant positive correlation between unit price and
total sales. This indicates that pricing strategies can directly influence revenue generation
without negatively affecting customer satisfaction. Interestingly, while the mean customer
satisfaction rating was below the target benchmark of 8, there was no significant correlation
between ratings and gross income, suggesting that customer satisfaction improvements alone
may not immediately boost profitability. Instead, they should be paired with robust pricing
and product strategies to optimize financial outcomes.

In terms of payment preferences, the study found no significant association between customer
type (member vs. walk-in) and payment methods. Both groups showed balanced usage of
digital and traditional payment options, highlighting a growing trend toward digital adoption.
This presents opportunities for the supermarket to expand its e-wallet offerings or collaborate
with digital payment platforms. Moreover, the average transaction value exceeded $300,
indicating strong overall performance. However, the wide range of transaction sizes points to
diverse shopping behaviors, suggesting opportunities to implement segment-specific
marketing strategies to cater to both high-spending and small-basket customers.

5 Limitations of the Study

Despite its contributions, the study faced several limitations. The data for key variables failed
normality tests, restricting the use of more powerful parametric statistical methods and
potentially limiting the depth of insights. Additionally, the analysis focused on a single
dataset from an online repository, which might not fully represent real-world supermarket
operations. The cross-sectional nature of the data provided only a snapshot of performance,
precluding the identification of trends or seasonal effects, which are crucial for long-term
strategic planning. Furthermore, the absence of more detailed demographic information, such
as age, income, or household size, constrained the ability to perform advanced customer
segmentation.

The study primarily relied on quantitative metrics, leaving qualitative aspects, such as
customer perceptions or preferences, unexplored. This limitation reduces the ability to gain a
holistic understanding of customer behavior. Moreover, external factors like promotions,
seasonality, and competitor actions were not considered, although they are likely to
significantly influence sales and customer satisfaction.

Implications for Future Research

Future studies can address these limitations by incorporating longitudinal data to analyze
trends and seasonal effects, as well as integrating additional variables like promotional
campaigns and competitor pricing. Expanding the dataset to include diverse supermarket
chains and geographic regions would improve generalizability. Collecting detailed
demographic information and qualitative feedback through surveys or interviews would
provide a more nuanced understanding of customer preferences. By addressing these areas,
future research can develop more robust insights to enhance decision-making in the
supermarket industry, ultimately driving customer satisfaction and profitability.

6 REFERENCES
 Babin, B. J., Darden, W. R., & Griffin, M. (2005). Work and/or fun: Measuring
hedonic and utilitarian shopping value. Journal of Consumer Research, 20(4), 644–
656. Retrieved from [Link]
 Choi, J., Lee, J., & Lee, J. (2020). The impact of digital and contactless payment
systems on consumer behaviour: A study of young customers in retail. Journal of
Retailing and Consumer Services, 54, 101924. Retrieved from:
[Link]
 Fisher, J., McNamara, M., & Robinson, L. (2017). Regional disparities in retail sales:
The influence of foot traffic, competition, and socio-economic factors. Journal of
Retailing and Consumer Services, 36, 156–167. Retrieved from
[Link]
 Kotler, P., & Keller, K. L. (2016). Marketing management (15th ed.). Pearson
Education. Retrieved from [Link]
management/P100000670185
 Nagle, T. T., & Holden, R. K. (2016). The strategy and tactics of pricing: A guide to
growing more profitably (5th ed.). Pearson. Retrieved from
[Link]
 Kaggle. (n.d.). super market sales dataset. Retrieved from:
[Link]