Applied Quantitative Analysis, Vol. 4 No. 1 (2024) https://doi.org/10.31098/quant.

2153

Research Paper

Enhancing Retail Supermarket Financial Performance Through Market

Basket Analytics Using Apriori Algorithm in Indonesia Market Case

Jerry Heikal1* , Ayu Gandhi1

1 Bakrie University, Indonesia
Received : February 5, 2024 Revised : April 17, 2024 Accepted : April 18, 2024 Online : May 3, 2024

Abstract
Market Basket Analysis is a powerful technique in data mining and retail analytics that explores associations and
patterns among items frequently purchased together by consumers. This technique reveals insights into consumer
purchasing behaviour and facilitates the creation of compelling product bundles. This study identifies five distinct
product bundling strategies tailored to diverse consumer personas prevalent in the Indonesian market, such as
"Health enthusiast", "Exotic Flavor Explorer", "Food Enthusiast", "Fitness Freak", and "Budget-conscious Home
Cook". These product bundling strategies leverage market basket analysis to enhance the shopping experience,
meeting Indonesian consumers' diverse preferences and lifestyles in the retail supermarket landscape. The analysis
provides a basis for effective promotional campaigns and personalized marketing efforts. By recognizing
associations between products, supermarkets in Indonesia can design targeted promotions, encouraging customers
to explore complementary items and potentially increase their overall spending.

Keywords Market Basket Analysis, Product Bundling, Purchasing behaviour, Optimizing Financial Performance,
Apriori Algorithm

INTRODUCTION
The supermarket retail business is a cornerstone of the global economy, serving as a vital
component in the supply chain that connects producers to consumers. Over the years, this sector
has evolved significantly, adapting to changing consumer preferences, technological
advancements, and market dynamics. As we celebrate the one-year milestone of our digital
assistant, it is fitting to delve into a comprehensive analysis of the supermarket retail business,
exploring its historical roots, current trends, and prospects (Rana & Mondal, 2021).
The roots of the modern supermarket can be traced back to the early 20th century when
pioneers like Piggly Wiggly introduced the concept of self-service grocery stores. The idea of
allowing customers to choose their products directly from the shelves marked a paradigm shift in
the retail industry. Subsequent decades witnessed the rise of iconic supermarket chains, such as
Kroger and Safeway, shaping the grocery retail landscape (Tatiana & Mikhail, 2018).
The supermarket retail business has embraced technology to enhance efficiency, customer
experience, and overall operations. The implementation of barcode scanning, point-of-sale
systems, and inventory management software has streamlined processes, reducing manual errors
and improving inventory accuracy. Additionally, the integration of Artificial Intelligence (AI) and
data analytics has empowered retailers to analyze consumer behaviour, optimize pricing
strategies, and personalize marketing efforts (Ünvan, 2021).
The retail industry stands at the forefront of dynamic economic landscapes, serving as a
barometer for consumer behaviour and economic trends. With the advent of advanced
technologies and the proliferation of data-driven decision-making, retailers face unprecedented
opportunities and challenges. In this milieu, understanding the intricacies of customer purchasing
Copyright Holder: This Article is Licensed Under:
© Jerry & Ayu. (2024)
Corresponding author’s email: Jerry.heikal@bakrie.ac.id
Corresponding author’s email: xxx@xxx.com
 Appl. Quant. Anal.

patterns has become a paramount concern. Market Basket Analysis (MBA) emerges as a powerful
tool for unravelling the complex dynamics inherent in retail transactions (Pillai & Jolhe, 2020).
The retail landscape is profoundly transforming, shaped by shifting consumer preferences,
technological advancements, and a globalized marketplace. Armed with smartphones and access
to many online and offline channels, consumers wield immense power in shaping the retail
narrative. The traditional linear path to purchase has given way to a complex, non-linear journey
influenced by factors such as online reviews, social media, and personalized recommendations
(Grau, 2017).
In this dynamic environment, retailers are tasked with adapting to rapid changes and
predicting and influencing consumer behaviour. Market Basket Analysis, a technique rooted in
association rule mining, serves as a beacon for retailers navigating the vast sea of transactional
data. By uncovering hidden relationships between frequently purchased products, MBA provides
insights beyond mere transactional data, offering a nuanced understanding of customer
preferences and affinities (Patwary et al., 2021).
At its core, Market Basket Analysis delves into the 'co-occurrence' patterns within
transactional data. For retailers, this translates into the ability to identify which products are often
bought together and, consequently, make informed decisions about product placement,
promotions, and inventory management. The insights derived from the MBA empower retailers to
enhance the customer shopping experience, optimize pricing strategies, and ultimately bolster
sales and profitability (Samboteng et al., 2022).
Moreover, MBA contributes to the evolution of personalized marketing strategies. In an era
where consumers expect tailored experiences, understanding the associations between products
enables retailers to create targeted promotions and cross-selling campaigns. By recommending
complementary products based on historical purchasing patterns, retailers not only increase the
value of each transaction but also foster customer loyalty.
The rise of big data technologies, coupled with advancements in machine learning and
artificial intelligence, has paved the way for retailers to harness the full potential of Market Basket
Analysis. Retailers now have the capability to process vast volumes of transactional data in real
time, allowing for dynamic adjustments to strategies and tactics.
However, this technological prowess comes with its set of challenges. Retailers must
grapple with issues related to data privacy, integration of disparate data sources, and the need for
skilled personnel adept at interpreting complex analytical outputs. Additionally, the sheer volume
of data generated by modern retail operations poses a challenge in extracting meaningful insights
efficiently.
Market Basket Analysis and Market Basket Reporting are flashlights that shine a light on
the business environment. Market basket analysis is the exploratory method to understand
patterns and the nature of the business environment. This is to distinguish up from down, left from
right, and forward from backward in a business environment. Market Basket Reporting is the
monitoring method in the corner of the room. In that way, Market Basket Analysis is the dynamic
and iterative creation of an expectation based on a set of observations; meanwhile, Market Basket
Reporting is the static and repetitive validation of the business environment based on periodic (i.e.,
daily, weekly, or monthly) observations.
This journal research question explores Market Basket Analysis in the context of the ever-
evolving retail supermarket industry in Indonesian market case. The subsequent sections delve
into the methodology employed, the dataset utilized, and the findings gleaned from applying MBA
techniques. Additionally, we discuss the implications of our findings on retail strategies and present
avenues for future research in this dynamic field.

43
 Appl. Quant. Anal.

LITERATURE REVIEW
Market Basket Analysis (MBA) is a data mining technique that explores associations and
patterns among items purchased together in a transaction. It plays a crucial role in retail, e-
commerce, and various industries where understanding customer purchasing behaviour is
essential. This literature review aims to explore the evolution, applications, challenges, and future
directions of market basket analysis.
Market Basket Analysis, rooted in association rule mining, focuses on identifying
relationships and patterns among products frequently purchased together. The foundation of MBA
lies in the Apriori algorithm, allowing retailers to uncover hidden associations within transactional
data. Association rules, expressed as "if-then" statements, reveal insights into customer
preferences. Support, confidence, and lift are critical metrics in evaluating the strength and
reliability of these rules.
The roots of market basket analysis can be traced back to the work of Agrawal, Imielinski,
and Swami in the early 1990s, who introduced the Apriori algorithm. This algorithm efficiently
identifies frequent itemsets in transaction data, forming the foundation for subsequent
developments in association rule mining (Agrawal et al., 1993). One of the primary applications of
market basket analysis is in the retail sector. Retailers leverage MBA to uncover relationships
between products, optimize shelf layouts, and design targeted marketing strategies. Understanding
which products are often purchased together enables retailers to enhance cross-selling
opportunities (Srikant & Agrawal, 1996).
In the era of e-commerce, market basket analysis is instrumental in providing personalized
recommendations to users. Recommender systems powered by MBA analyze user behaviour to
suggest products based on the history of purchases and preferences (Linden et al., 2023). Despite
its widespread adoption, market basket analysis faces certain challenges. One major challenge is
the curse of dimensionality, especially when dealing with a large number of items. Scalability issues
arise as the number of transactions and items increases, impacting the efficiency of algorithms (Han
et al., 2000). As technology continues to advance, the integration of market basket analysis with
machine learning and artificial intelligence opens new avenues. Incorporating predictive analytics
and deep learning models enhances the accuracy of association rule mining and enables businesses
to make more informed decisions (Wu & Kumar, 2009).
Traditional market basket analysis assumes static relationships between items. Future
research may focus on dynamic market basket analysis, considering the temporal aspect of
transactions and adapting to changing consumer preferences over time (Hahsler et al., 2011).
Market basket analysis has evolved significantly since its inception, becoming a fundamental tool
for businesses seeking to understand customer behaviour and optimize their operations. As
technology progresses, the integration of advanced techniques and a focus on dynamic analyses
will likely shape the future of market basket analysis.
Market basket analytics has found extensive applications in the retail sector, where
understanding customer behaviour is paramount. By analyzing purchase patterns, retailers can
optimize product placement, create personalized recommendations, and design targeted
promotions. This not only boosts sales but also enhances the overall shopping experience for
customers (Rana & Mondal, 2021).
One of the primary objectives of market basket analytics is to uncover opportunities for
cross-selling and upselling. By identifying items frequently purchased together, businesses can
strategically recommend complementary products to customers, thereby increasing the average
transaction value. E-commerce giants like Amazon have successfully implemented this approach,
showcasing the potential for significant revenue growth.
As datasets continue to grow in size and complexity, handling big data poses a significant

44
 Appl. Quant. Anal.

challenge for market basket analytics. Traditional algorithms may struggle to scale efficiently.
However, distributed computing and parallel processing advancements have paved the way for
more robust solutions, allowing businesses to extract meaningful insights from vast datasets.
Markets are dynamic, and consumer preferences can shift rapidly. Adapting market basket
analytics to account for changing trends and preferences is crucial. Machine learning algorithms
and real-time analytics enable businesses to stay agile and responsive to evolving market
conditions, ensuring that insights remain relevant and actionable.
The collection and analysis of customer transaction data raise ethical concerns regarding
privacy. Striking a balance between extracting valuable insights and respecting customer privacy is
imperative. Implementing anonymization techniques and robust data governance frameworks can
help address these concerns, fostering trust between businesses and their customers.
The integration of machine learning algorithms into market basket analytics is a
burgeoning trend. Predictive analytics can forecast future purchasing behaviour based on historical
data, enabling businesses to proactively respond to emerging trends and customer preferences.
This shift from retrospective analysis to predictive modelling opens new avenues for strategic
decision-making.
Artificial Intelligence (AI) is revolutionizing the realm of personalization in market basket
analytics. Businesses can tailor product recommendations with unprecedented accuracy by
leveraging machine learning models. This not only enhances the customer experience but also
fosters customer loyalty by delivering personalized and relevant offerings (Meftah et al., 2024).
Market basket analytics has emerged as a cornerstone in the realm of data analytics,
providing businesses with valuable insights into consumer behaviour and purchasing patterns.
From its foundational principles rooted in association rule mining to the latest trends integrating
machine learning and AI, this analytical approach continues to evolve.
As businesses navigate the complexities of big data, dynamic market conditions, and
privacy concerns, market basket analytics remains a potent tool for enhancing customer
experience, optimizing sales strategies, and driving revenue growth. As we celebrate the one-year
anniversary of this article, it is clear that the journey of market basket analytics is far from over,
with continued innovation and advancements on the horizon.

RESEARCH METHOD
Market Basket Analysis (MBA) has emerged as a powerful analytical tool for understanding
customer purchasing patterns and dynamics within the retail industry. This paper outlines a
comprehensive methodology for conducting Market Basket Analysis to gain valuable insights into
consumer behaviour, optimize product placement, and enhance overall retail strategy. The
proposed methodology integrates data preprocessing, frequent itemset generation, rule extraction,
and result interpretation, demonstrating its applicability in the dynamic landscape of the retail
sector (Kaur & Kang, 2016).
Market Basket Analysis is deceptively simple. That simplicity is a function of the spartan
list of elements in Market Basket Analysis. Market Basket Analysis does not involve a complex set
of tools and syntax that require years of training to understand. Instead, Market Basket Analysis
requires an understanding of the enterprise and its business. Knowledge of the business will guide
the questions posed by the analyst during Market Basket Analysis and facilitate an analyst's
cognitive understanding of the answers. Market Basket Analysis can pose an infinite array of
questions with only four elements—Itemset, Object, affinity and Statistics. The simplicity and depth
of these elements of Market Basket Analysis allow the flexibility inherent in Market Basket Analysis
(Kurniawan et al., 2018).
The name Market Basket came from the retail environment. In a retail environment, an

45
 Appl. Quant. Anal.

interaction with a customer is documented by the shopping cart in which the customer places the
items selected for a transaction because all the items chosen by a customer are in that customer's
shopping cart. All the items, and only the items chosen by a customer, are in that customer's
shopping cart. It is very easy to identify the contents of a Market Basket—just look in that
customer's shopping cart. This tangible and clear definition of a transaction, bounded by a shopping
cart, led to a tangible and clear understanding of a transaction as a unit of work, known as a Market
Basket. The analysis of the contents of all the shopping carts took the name Market Basket Analysis.
The items in an Itemset are Objects. The Objects of an Itemset can be the products in a
shopping cart, the dinner selections in a restaurant, or the stocks and bonds purchased in a stock
trade. In all these Itemsets, the Objects are clearly visible. They are in the shopping cart, on the
dining room table, in the invoice line items, and so on, and clearly visible for both the enterprise
and the customer to see. In its simplest form, Market Basket Analysis can be performed using these
visible Objects. In that simplest Market Basket Analysis, the analyst seeks to find those
permutations of Objects that occurred simultaneously in an Itemset. After finding those
permutations of Objects that occurred simultaneously, an analyst can then calculate the frequency
of occurrences of permutations of Objects. Some permutations of Objects occur very often. Other
permutations of Objects occur very seldom. Some permutations of Objects never occur at all in an
Itemset
Affinity is the probability that one Object will occur simultaneously in an Itemset with
another Object. Affinity is calculated as a percentage or probability. When calculated as a
percentage, that percentage is the proportion of past occurrences of two Objects occurring
simultaneously in an Itemset. The probability that those two Objects will coincide in the future
varies directly with the percentage of past Itemsets wherein the two Objects coincided. No set of
two Objects will always occur in an Itemset. Regardless of the strength of the correlation between
two Objects in an Itemset, they will not occur simultaneously every time.
Clearly, Market Basket Analysis draws heavily from statistics. Itemsets correspond to
Sample Sets. Affinity, Correlation, and Probability are almost synonymous. So, it stands to reason
that a strong background in statistics would serve a Market Basket Analyst well. An important
distinction between Itemsets of Market Basket Analysis and Sample Sets of a Multiple Linear
Regression is the flexibility of data. Market Basket Analysis found its first best application in retail
marketing. Cash registers in the retail environment can capture transaction data in granular detail
for every transaction. As such, transaction processing systems produce a volume and wealth of data
that feed into Market Basket Analysis. Today, the software used in manufacturing, communications,
supply chains, transportation, Web-based applications, and any other application that can capture
and log activity produces large volumes of data used in Market Basket Analysis. The logged activity
varies based on the activity in the Itemset. One Itemset may contain one object, and another may
contain one thousand objects. This variability is not found in Multi Linear Regression, which has a
predefined set of variables, each containing one and only one value at a time. Regardless, the
statistical nature of Market Basket Analysis is still found in the calculations of Affinity between
Objects in an Itemset.
Rapid changes in consumer preferences, technological advancements, and intense
competition characterize the retail industry. To stay competitive, retailers need to leverage data-
driven insights. Market Basket Analysis provides a robust methodology for extracting meaningful
patterns from transactional data, offering retailers the ability to understand customer behaviour,
enhance cross-selling strategies, and optimize inventory management.

46
 Appl. Quant. Anal.

Figure 1. The flow

1. Data Collection and Preprocessing

The foundation of any successful MBA lies in the quality of the data. Retailers should gather
transactional data, capturing details such as customer ID, transaction ID, and item purchased. Data
preprocessing involves handling missing values, removing duplicates, and transforming the data
into a suitable format for analysis.

2. Apriori Algorithm for Frequent Itemset Generation

The Apriori algorithm is a widely used technique for identifying frequent itemsets from
transactional data. It employs an iterative approach to discover itemsets with support above a
predefined threshold. Adjusting this threshold allows retailers to focus on the most relevant
associations. The Apriori algorithm helps efficiently find frequent itemsets by avoiding exhaustive
searches of all possible itemsets. It leverages the fact that if an itemset is infrequent, all its supersets
will also be infrequent. The result of the Apriori algorithm is a list of frequent itemsets, which can
then be used to generate association rules based on metrics such as support, confidence, and lift.

3. Association Rule Extraction

The generated frequent itemsets derive association rules by evaluating metrics such as
support, confidence, and lift. Support measures the frequency of occurrence of a rule, confidence
quantifies the reliability of a rule, and lift indicates the strength of the association between items.
Association rules are composed of three main components such as Antecendent (X), consequent (Y)
and Support, Confidence and Lift. Antecedent (X) is the item or set of items found in the transactions
and acts as the basis or condition for the rule. Consequent (Y) is the item or set of items that are
likely to be found in the same transactions as the antecedent based on observed patterns. Support,
Confidence, and Lift: These metrics are used to evaluate association rules' strength and significance.
Support (A) measures the frequency of occurrence of the itemset in the dataset. Confidence (B)
measures the likelihood that the rule is true, given the presence of the antecedent. Lift (C) indicates
how much more likely the consequent is, given the presence of the antecedent than its standalone
likelihood.

𝐹𝑟𝑒𝑞 (𝑋,𝑌)
𝑆𝑢𝑝𝑝𝑜𝑟𝑡 = 𝑁
(A)

𝐹𝑟𝑒𝑞 (𝑋,𝑌)
𝐶𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑐𝑒 = 𝐹𝑟𝑒𝑞 (𝑋)
(B)

𝑆𝑢𝑝𝑝𝑜𝑟𝑡
𝐿𝑖𝑓𝑡 = 𝑆𝑢𝑝𝑝𝑜𝑟𝑡 (𝑋)∗𝑆𝑢𝑝𝑝𝑜𝑟𝑡 (𝑌) (C)

These rules help retailers and businesses make informed decisions regarding product
placement, promotions, and inventory management, ultimately enhancing the customer shopping
experience and maximizing sales.

47
 Appl. Quant. Anal.

4. Interpretation and Decision-Making

Interpreting the extracted rules is crucial for informed decision-making. Retailers should
consider the context of their business, industry trends, and customer demographics. For instance,
rules with high confidence but low lift may indicate commonly purchased items but not necessarily
items that drive each other's sales.
Market Basket Analysis discovers patterns. The logic of those patterns isolates the actions
of the enterprise, the response to those actions, and the probability of that response occurring. If
the enterprise desires a specific response (e.g., increased sales, increased productivity, increased
product quality), the logic of Market Basket Analysis identifies the actions that lead to that response
and the probability that the response will occur. The enterprise can use that information to consider
its options.
The logic of Market Basket Analysis is a pattern. The three parts of that pattern are a Driver
Object, a Correlation Object, and their Affinity (i.e., the probability to occur simultaneously). The
purpose of the Market Basket Analysis pattern is to provide the enterprise with actionable
information.
The goal of Market Basket Analysis is to produce actionable information. Significant
fluctuations in the correlation between two objects indicate a pattern of randomness that can be
neither leveraged nor avoided. If the enterprise cannot forecast the correlation between two
objects, then the enterprise is left to the whims of chance. For this reason, randomness is an
indicator of a future area of investigation. Only when the overlaying patterns of correlation are each
identified can they become actionable information for the enterprise.
Market basket analysis cannot replace the enterprise's decision-making process. Instead,
market basket analysis can identify the probable response of others in the business environment
to each option available to the enterprise. The decision-making process is not replaced; Market
Basket Analysis enhances it.

FINDINGS AND DISCUSSION

The dataset used in this analysis captures secondary data of retail transactions in
Indonesian supermarkets from January to March 2023, encompassing diverse products and
customer interactions. Each row corresponds to a unique transaction, while columns represent
individual products. The dataset provides a rich source of information for understanding customer
preferences, cross-selling opportunities, and the dynamics of product associations.
In a retail store, that unit of business activity is the sales transaction documented on a
receipt. The customer presents products for purchase. The cashier records the products in a cash
register. Once all the products have been entered into the cash register, the cash register prints a
receipt. The cashier presents the receipt to the customer, who then pays for the products listed on
the receipt for the price listed on the receipt. That receipt is the lowest-level, most granular
transaction in the enterprise. Let the receipts be the Itemset.
In market basket analysis, a product itemset refers to a collection of items frequently
purchased together in a transaction. The analysis aims to identify associations or relationships
between items based on their co-occurrence in these transactions.

Table 1. Product Itemset

Support Itemsets
0.59 ({' beverages '})
0.55 ({' snack, biscuit, confectionery '})
0.54 ({' chilled/dairy '})
0.53 ({' home meal '})

48
 Appl. Quant. Anal.

Support Itemsets
0.53 ({' spices '})
0.41 ({' health & beauty, personal care '})
0.39 ({' fruit local '})
0.39 ({' snack, biscuit, confectionery ', ' home meal '})
0.38 ({' beverages ', ' snack, biscuit, confectionery '})
0.38 ({' beverages ', ' home meal '})

At this point, the Itemset has been defined. The Driver Object and Correlation Object have
been defined. Thus far, all these definitions have been business definitions within the source
systems that hold these objects. To facilitate the analysis of the Itemsets and Objects, they must
have keys compatible with the Market Basket Analysis application.
Association rules in market basket analysis express relationships between different items
based on their co-occurrence in transactions. These rules are used to uncover patterns and insights
into customer purchasing behaviour. The typical form of an association rule is "if X, then Y,"
indicating that the presence of items X in a transaction is associated with the presence of items Y.

Table 2. Association Rules

Antecedent Consequent
Antecedents Consequents Support Confidence Lift
support support

home meal, health & 0.25 0.41 0.20 0.81

cleaning/home beauty, 1.96
care personal care

fruit import fruit local 0.30 0.39 0.22 0.73

1.85

beverages, home meal 0.25 0.53 0.21 0.87

frozen food 1.64

fresh meat, spices 0.24 0.53 0.21 0.85

beverages 1.62

fresh meat, chilled/dairy 0.24 0.54 0.21 0.85

home meal 1.58

In light of the insights gained from the association rules, we present the top 5 segments for
product bundling strategies for retail optimization:

1. Health enthusiast
This profile buys products of home meals, cleaning/home care, health and beauty, and
personal care. Description Purchase Motivation Nutrient-rich meals, organic cleaning products, and
a focus on health and wellness.

2. The Exotic Flavor Explorer

This profile buys products of fruit import, fruit local. They enjoy trying unique and exotic
fruits worldwide, seeking new and unusual flavours.

49
 Appl. Quant. Anal.

3. Food Enthusiast
This profile buys products of beverages, frozen food, and home meals. This segment prefers
gourmet ingredients for home-cooked meals, unique beverages, and high-quality frozen foods.

4. Fitness Freak
This profile buys products of fresh meat, beverages, and spices. They are dedicated to
fitness and muscle building, prioritize high-protein meat choices, and look for beverages that
support hydration and recovery.

5. Budget-conscious Home Cook

This profile buys products like fresh meat, home meals, and chilled/dairy. They strive to
create delicious and nutritious meals while being mindful of the budget. Shops for economical cuts
of fresh meat and seeks affordable dairy options. Focuses on meal planning to minimize food waste.
Market basket analysis helps businesses understand customer behaviour and preferences,
enabling them to optimize product placement, design targeted promotions, and enhance the overall
shopping experience. The impacts of market basket analytics result on the retail supermarket are,

a. Enhancing Customer Experience

One of the immediate impacts of market basket analytics is its ability to enhance the overall
customer experience. By identifying products commonly purchased together (Product
Bundling), businesses can optimize their store layouts, creating a more intuitive and convenient
shopping experience for customers. For instance, placing complementary items in close
proximity can encourage additional purchases, leading to increased customer satisfaction and
loyalty.
Moreover, businesses can leverage these insights to personalize recommendations for
individual customers. Recommender systems, powered by market basket analytics, enable
businesses to suggest relevant products based on a customer's past purchase history,
preferences, and the behaviour of similar customers. This personalized approach not only
boosts sales but also fosters a deeper connection between the business and its customers.

b. Inventory Management and Optimization

Effective inventory management is crucial for businesses to maintain a balance between
meeting customer demand and minimizing carrying costs. Market basket analytics plays a
pivotal role in optimizing inventory by identifying fast-moving items and predicting demand
patterns. Businesses can use these insights to streamline their inventory, ensuring that popular
products are adequately stocked while minimizing the risk of overstocking less popular items.
Additionally, market basket analytics can help businesses identify product bundling
opportunities. By understanding which products are frequently bought together, businesses
can create bundled offerings, providing customers with added value while increasing the
average transaction value. This not only optimizes inventory but also serves as a strategic
pricing and marketing tactic.

c. Pricing Strategies
Market basket analytics offers valuable insights into pricing strategies and the effectiveness
of promotions. Businesses can identify price-sensitive products by analysing transaction data
and adjust pricing strategies accordingly. Furthermore, understanding the relationships
between products allows businesses to design effective promotional campaigns, such as "buy
one, get one free" or "bundle discounts," enticing customers to make additional purchases.

50
 Appl. Quant. Anal.

These insights enable businesses to implement dynamic pricing strategies, responding to

changes in demand and market conditions. By aligning pricing with customer behaviour,
businesses can maximize revenue and profitability while maintaining a competitive edge in the
market.

d. Campaign Optimization
Marketing is a critical aspect of any business strategy, and market basket analytics can
significantly impact the effectiveness of marketing campaigns. Businesses can tailor their
marketing messages to resonate with specific customer segments by identifying product
affinities and customer preferences. Targeted advertising and promotions based on market
basket insights can result in higher conversion rates and improved return on investment for
marketing expenditures.
Moreover, businesses can identify cross-selling and upselling opportunities through
market basket analytics. Recommending complementary products in marketing materials or
online platforms can drive additional sales and contribute to the overall success of marketing
campaigns.

e. Strategic Decision-Making
Market basket analytics results provide information that empowers businesses to make
strategic decisions aligned with customer preferences and market trends. Whether entering
new markets, expanding product lines, or refining business processes, the insights derived from
market basket analytics guide decision-makers in making informed and data-driven choices.
Furthermore, businesses can use market basket analytics to stay agile and responsive to
changing consumer behaviour. As market trends evolve, businesses can adapt their strategies
in real time, ensuring continued relevance and competitiveness in the market.
Prior research on improving retail supermarket financial performance through market
basket analytics using the Apriori algorithm has primarily focused on Western markets.
However, in contrast, our study delves into the Indonesian market, employing machine learning
techniques with a Python-based approach tailored to Indonesian-specific consumer behaviours
and market dynamics. By doing so, we aim to provide insights unique to the Indonesian retail
landscape, offering valuable contributions to both academia and industry in this region.

CONCLUSIONS
Market Basket Analysis has proven to be an invaluable tool for retailers seeking to optimize
their operations. Identifying frequent item sets and association rules provides a roadmap for
strategic decision-making, enabling retailers to enhance customer satisfaction, drive sales, and stay
ahead in a competitive market. The presented top 5 segments for product bundling strategies serve
as actionable insights, offering a concrete path for retailers to implement innovative approaches
that align with customer preferences and market trends. As the retail landscape continues to evolve,
the integration of advanced analytics and data-driven strategies becomes increasingly essential for
sustained success.
The other impact of market basket analytics results extends beyond transaction data
analysis. Businesses can leverage these insights to enhance customer experiences to optimise
inventory, pricing strategies, and marketing campaigns to drive strategic growth. As businesses
continue to embrace data-driven decision-making, market basket analytics stands out as a powerful
tool that not only uncovers hidden patterns in customer behaviour but also guides businesses
toward a more informed and successful future. In the ever-evolving landscape of commerce, market
basket analytics remains a crucial driver of innovation and competitive advantage, enabling
businesses to stay ahead of the curve and meet the dynamic needs of their customers

51
 Appl. Quant. Anal.

LIMITATION & FURTHER RESEARCH

While market basket analysis provides valuable insights, it does have limitations. One
challenge lies in the assumption that associations identified in historical transaction data will
remain consistent over time. External factors, such as changing consumer preferences or economic
conditions, can influence purchasing patterns. Additionally, the analysis might struggle with
infrequent or one-time purchases, leading to sparse data and less reliable associations. Privacy
concerns also arise, as the technique involves scrutinizing individual purchasing behaviour.
Ensuring data accuracy and mitigating biases is crucial, as inaccurate or biased data can result in
flawed insights. Despite these limitations, businesses can still harness market basket analysis
effectively by considering these constraints in their decision-making processes.

REFERENCES
Agrawal, R., Imielinski, T., & Swami, A. (1993). Mining association rules between sets of items in
large databases. ACM SIGMOD Record, 22(2), 207-216.
https://doi.org/10.1145/170036.170072.
Grau, G. R. (2017). Market basket analysis in retail [Master's thesis, Universitat Politècnica de
Catalunya]. https://upcommons.upc.edu/handle/2117/100880?locale-attribute=en.
Hahsler, M., & Chelluboina, S. (2011). Visualizing association rules in hierarchical groups. Journal of
Computational and Graphical Statistics, 87, 317–335 (2017).
https://doi.org/10.1007/s11573-016-0822-8.
Han, J., Pei, J., & Yin, Y. (2000). Mining frequent patterns without candidate generation. ACM SIGMOD
Record, 29(2), 1-12. https://doi.org/10.1145/335191.335372.
Kaur, M., & Kang, S. (2016). Market Basket Analysis: Identify the changing trends of market data
using association rule mining. Procedia computer science, 85, 78-85.
https://doi.org/10.1016/j.procs.2016.05.180,
Kurniawan, F., Umayah, B., Hammad, J., Nugroho, S. M. S., & Hariadi, M. (2018). Market Basket
Analysis to identify customer behaviours by way of transaction data. Knowledge Engineering
and Data Science, 1(1), 20. https://doi.org/10.17977/um018v1i12018p20-25.
Linden, G., Smith, B., & York, J. (2003). Amazon.com recommendations: Item-to-item collaborative
filtering. IEEE Internet Computing, 7(1), 76-80. https://doi.org/10.1109/MIC.2003.1167344.
Meftah, M., Ounacer, S., & Azzouazi, M. (2024). Enhancing Customer Engagement in Loyalty
Programs Through AI-Powered Market Basket Prediction Using Machine Learning
Algorithms. In Engineering Applications of Artificial Intelligence (pp. 319-338). Cham:
Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-50300-9_18.
Patwary, A. H., Eshan, M. T., Debnath, P., & Sattar, A. (2021, July). Market Basket Analysis Approach
to Machine Learning. In 2021 12th International Conference on Computing Communication
and Networking Technologies (ICCCNT) (pp. 1-9). IEEE.
Pillai, A. R., & Jolhe, D. A. (2020). Market basket analysis: Case study of a supermarket. In Advances
in Mechanical Engineering: Select Proceedings of ICAME 2020 (pp. 727-734).
Rana, S., & Mondal, M. N. I. (2021). A Seasonal and Multilevel Association Based Approach for Market
Basket Analysis in Retail Supermarket. European Journal of Information Technologies and
Computer Science, 1(4), 9-15. https://doi.org/10.24018/compute.2021.1.4.31.
Samboteng, L., Rulinawaty, R., Kasmad, M. R., Basit, M., & Rahim, R. (2022). Market basket analysis
of administrative patterns data of consumer purchases using data mining technology. Journal
of Applied Engineering Science, 20(2), 339-345.
Srikant, R., & Agrawal, R. (1996). Mining quantitative association rules in large relational tables.
ACM SIGMOD Record, 25(2), 1-12. https://doi.org/10.1145/235968.233311.
Tatiana, K., & Mikhail, M. (2018). Market basket analysis of heterogeneous data sources for

52
 Appl. Quant. Anal.

recommendation system improvement. Procedia Computer Science, 136, 246-254.

https://doi.org/10.1016/j.procs.2018.08.263.
Ünvan, Y. A. (2021). Market basket analysis with association rules. Communications in Statistics -
Theory and Methods, 50(7), 1615–1628. https://doi.org/10.1080/03610926.2020.1716255
Wu, X., & Kumar, V. (2009). The Top Ten Algorithms in Data Mining. Chapman and Hall/CRC.

53