Academia.eduAcademia.edu

Outline

Title
Abstract
Figures
Introduction
Implementation A. Data Cleaning
Exploratory Data Analysis
References
All Topics
Computer Science

Customer Behavior Analysis using Machine Learning

Profile image of Deepika RoyDeepika Roy

2021, International Journal for Research in Applied Science and Engineering Technology

https://doi.org/10.22214/IJRASET.2021.35180
visibility

…

description

6 pages

Abstract

RFM (Recency, Frequency, Monetary) investigation is a demonstrated showcasing model for conduct based client division. It groups clients dependent on their exchange history – how as of late, how frequently and what amount they buy.RFM helps partition clients into different classes or groups to distinguish clients who will react to advancements and how. This RFM examination depends on a blend of three boundaries. For instance, we can say that individuals who spend the most on items are our best clients. A large portion of us coincide and think about something very similar. In any case, Imagine a scenario in which they were bought just a single time. Or on the other hand an extremely quiet past? Consider the possibility that they are done utilizing our item. would they be able to in any case be viewed as your best clients? Most likely not. Making a decision about client esteem from only one perspective will give you a mistaken report of your client base and their lifetime. That is the...

Figures (3)
A framework design is the model that characterizes the construction, conduct and more perspectives on the framework. It contains framework parts and the framework created that will cooperate to carry out the general framework. In the Machine learning design , the crude information goes through numerous means prior to giving the yield. The entire information is screened and irregularities and indistinguishable information is eliminated from the informational index bringing about the yield with complete precision.
A framework design is the model that characterizes the construction, conduct and more perspectives on the framework. It contains framework parts and the framework created that will cooperate to carry out the general framework. In the Machine learning design , the crude information goes through numerous means prior to giving the yield. The entire information is screened and irregularities and indistinguishable information is eliminated from the informational index bringing about the yield with complete precision.
TABLE-1 Product Categories In measurements , exploratory information examination (EDA) is a way to deal with breaking down informational collections t sum up their primary attributes, regularly with visual strategies. A factual model can be utilized or not, yet basically EDA is fo seeing what the information can advise us past the formal displaying or speculation. Exploratory information examination wa advanced by John Tukey to urge analysts to investigate the information, and conceivably figure speculations that could prompt nev information assortment and trials. EDA is unique in relation to starting information examination (IDA), which centers all the mor barely around checking suppositions needed for model fitting and speculation testing, and taking care of missing qualities an making changes of factors as needelInsights about the purchase behavior.
TABLE-1 Product Categories In measurements , exploratory information examination (EDA) is a way to deal with breaking down informational collections t sum up their primary attributes, regularly with visual strategies. A factual model can be utilized or not, yet basically EDA is fo seeing what the information can advise us past the formal displaying or speculation. Exploratory information examination wa advanced by John Tukey to urge analysts to investigate the information, and conceivably figure speculations that could prompt nev information assortment and trials. EDA is unique in relation to starting information examination (IDA), which centers all the mor barely around checking suppositions needed for model fitting and speculation testing, and taking care of missing qualities an making changes of factors as needelInsights about the purchase behavior.
— RT ee N. With the ML Algorithms like grouping calculations and comparable item suggestions utilized, the arrangement was constructed easily giving advantages to the steadfast clients. These techniques were recognized easily utilizing the AI strategies in these applications. This methodology can be utilized in any retail market deals information where understanding client prerequisite also, building procedures that would increment deals to the organization and furthermore fulfill the clients that should purchase the items by giving offers, limits and integral or combos to the clients. B. Clusters Formed After RFM Analysis The authors would like to acknowledge the support of the Chairman, Director and Head of the Department, Department of Computer Science and Engineering, CMR Technical Campus, Medchal, Hyderabad, Telangana, for their encouragement to the authors. REFERENCES
— RT ee N. With the ML Algorithms like grouping calculations and comparable item suggestions utilized, the arrangement was constructed easily giving advantages to the steadfast clients. These techniques were recognized easily utilizing the AI strategies in these applications. This methodology can be utilized in any retail market deals information where understanding client prerequisite also, building procedures that would increment deals to the organization and furthermore fulfill the clients that should purchase the items by giving offers, limits and integral or combos to the clients. B. Clusters Formed After RFM Analysis The authors would like to acknowledge the support of the Chairman, Director and Head of the Department, Department of Computer Science and Engineering, CMR Technical Campus, Medchal, Hyderabad, Telangana, for their encouragement to the authors. REFERENCES

Related papers

Customer clustering using RFM analysis
Vasilis Aggelis

2018 26th Signal Processing and Communications Applications Conference (SIU)

RFM (Recency, Frequency, Monetary) analysis is a method to identify high-response customers in marketing promotions, and to improve overall response rates, which is well known and is widely applied today. Less widely understood is the value of applying RFM scoring to a customer database and measuring customer profitability. RFM analysis is considered significant also for the banks and their specific units like e-banking. A customer who has visited an e-banking site Recently (R) and Frequently (F) and created a lot of Monetary Value (M) through payment and standing orders is very likely to visit and make payments again. After evaluation of the customer's behaviour using specific RFM criteria the RFM score is correlated to the bank interest, with a high RFM score being more beneficial to the bank currently as well as in the future. Data mining methods can be considered as tools enhancing the bank RFM analysis of the customers in total as well as specific groups like the users of e-banking.

View PDFchevron_right
Application of RFM Model on Customer Segmentation in Digital Marketing
Akazue Maureen

Deleted Journal, 2024

Customers play a pivotal role in the success of any business. The ability to attract and retain the right clientele, who consistently engage with a company's products and services, hinges on a thorough understanding of their purchasing behavior. Successful businesses tailor their offerings to meet the unique requirements and preferences of their customers. Utilizing marketing analysis tools, such as the RFM model, facilitates the segmentation of customers based on distinct parameters, enabling the identification of high-value customers through factors like purchase behavior. In addressing this critical aspect of customer-centric strategies, our research introduces an innovative early purchase prediction framework. This framework aims to assess the likelihood of a potential customer purchasing soon. Leveraging a combination of a decision tree classifier and a gradient-enhancing classification approach, we tackled the classification challenge using a dataset obtained from statistadata.com, focusing on digital market dynamics. Our findings reveal that the gradient-enhancing classifier model outperformed others, demonstrating an impressive accuracy of 93% and an AUC (Area Under the Curve) of 0.98. This research not only contributes to the advancement of predictive modeling in the realm of customer behavior but also underscores the practical application of such models in digital marketing strategies. As businesses navigate the intricacies of customer engagement, our proposed framework offers valuable insights to enhance decision-making processes and optimize marketing efforts.

View PDFchevron_right
The Implementation of RFM Analysis to Customer Profiling Using K-Means Clustering
Jefri Junifer Pangaribuan

Mathematical Modelling of Engineering Problems

The vast development of information technology causes an explosion in the amount of data, yet the data must be processed to obtain useful insights. The use of data is needed to study the needs, behavior, and customer's value which are meant to build better relationships or what is often referred to Customer Relationship Management (CRM). As the company grows, data is getting abundant and more difficult to interact directly with customers and problems such as marketing campaigns that are less effective can result in losses if not immediately addressed. Therefore, customer segmentation was carried out using recency, frequency, and monetary (RFM) as variables and K-Means clustering by determining the number of clusters using the elbow method and silhouette score. Based on the analysis results, there are three types of clusters, categorized as best customers, may not lost customers, and average customers.

View PDFchevron_right
DEVELOPING RFM MODEL FOR CUSTOMER SEGMENTATION IN RETAIL INDUSTRY
iaeme iaeme

iaeme

Many of the research tools used by sophisticated consumer marketing companies for new product development fall into the broad category referred to as "market segmentation." Recently some of these segmentation techniques have come under considerable fire. In this article we will design a mathematical model for the segmentation of retail customers. In this paper, we have segmented them based on recency, frequency and monetary (RFM) measures. Data required applying this method gathered from al the Retail Stores of Odisha which is established newly.

View PDFchevron_right
Customer Segmentation and Strategy Development Based on User Behavior Analysis, RFM Model and Data Mining Techniques: A Case Study
Mohammadreza Tavakoli

2018 IEEE 15th International Conference on e-Business Engineering (ICEBE), 2018

The RFM (Recency, Frequency and Monetary) model provides an effective analysis for decision makers in order to target their customers and develop appropriate marketing strategies according to their previous behaviors. Although the RFM model has been widely applied in various areas of marketing, its simplicity threatens its effectiveness since it does not consider the customers' relationship and changes in customers' behavior. In this paper, we propose an R+FM model which configures the segmentation according to the business changes and clusters customers using K-Means. We applied our model on Digikala company, the biggest E-Commerce in Middle East, and compared our model with the Digikala's previous RFM model which used Customer Quantile Method. Moreover, we built strategies for each segment and ran an SMS campaign according to those strategies. The results of the campaign showed that our Segmentation Model improved the number of purchase and average monetary of the baskets.

View PDFchevron_right
CRM: Tools to retrain the customer and identify Customer Purchasing Pattern
Sujo George

2020

In order to sustain in the market, a marketer must know the customer needs and wants and their purchasing pattern and the market trend. In this paper, we have focused on the marketer point of view, how he/ she must look towards the opportunities and competitors side by using Plan,Do and Evaluate.we have tried to cover all the aspects of the organization. Here I have taken the primary survey on an online platform like what are their common problems, offers, how often they visit and small analysis on the data. The Marketer must focus on long term goals by having good relations with the customers. Every single feedback and every complaint should be focused with due care and respect to deliver the customer best.

View PDFchevron_right
Estimating customer lifetime value based on RFM analysis of customer purchase behavior: Case study
Sarah Ashoori

Procedia Computer Science, 2011

Since the increased importance is placed on customer equity in today's business environment, many firms are focusing on the notion of customer loyalty and profitability to increasing market share. Building successful customer relationship management (CRM), a firm starts from identifying customers' true value and loyalty since customer value can provide basic information to deploy more targeted and personalized marketing. In this paper, customer lifetime value (CLV) is used to customer segmentation of a health and beauty company. Two approaches are used: in the first approach, RFM (Recency, Frequency, and Monetary) marketing analysis method is used in order to segmentation of customers and in the second approach, the proposed extended RFM analysis method with one additional parameter -called Count Item -is used. Comparing results of these approaches, shows that adding count Item as a new parameter to RFM method makes no difference to clustering result, so CLV is calculated based on weighted RFM method for each segment. The results of calculated CLV for different segments can be used to explain marketing and sales strategies by the company.

View PDFchevron_right
Comparison of customer response models
Ching Gu

Service Business, 2009

Segmentation of customers by likelihood of repeating business is a very important tool in marketing management. A number of approaches have been developed to support this activity. This article reviews basic recency, frequency, and monetary (RFM) methods on a set of data involving the sale of beef products. Variants of RFM are demonstrated. Classical data mining techniques of logistic regression, decision trees, and neural networks are also demonstrated. Results indicate a spectrum of tradeoffs. RFM methods are simpler, but less accurate. Considerations of balancing cell sizes as well as compressing data are examined. Both balancing expected cell densities as well as compressing RFM variables into a value function were found to provide more accurate models. Data mining algorithms were all found to provide a noticeable increase in predictive accuracy. Relative tradeoffs among these data mining algorithms in the context of customer segmentation are discussed.

View PDFchevron_right
Customer Behavior Analysis in E-Commerce using Machine Learning Approach : A Survey
International Journal of Scientific Research in Computer Science, Engineering and Information Technology IJSRCSEIT

International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 2023

These days, consumer behavior models are often founded on machine learning and the data mining of customer data. Forecasting client behavior is an unclear and tough endeavor. Thus, one has to use the appropriate strategy and strategy while constructing consumer behavior models. Once a prediction model has been built, it is difficult to manipulate it for the marketer to determine exactly what marketing actions to take for each customer or group of customers. This is because once the model has been built, it is impossible to change the variables that make up the model. While this formulation is very complicated, the majority of customer models are, in practice, rather straightforward. Because of this requirement, most customer behavior models neglect so many essential elements that the forecasts they provide are often not particularly trustworthy. The purpose of this study is to present different research work that has been done on the analysis of consumer behavior using various machine learning and data mining approaches.

View PDFchevron_right
IRJET- Customer Relationship Management Using Machine Learning
IRJET Journal

IRJET, 2021

At the present time, customers have shown more interest in the quality of service that the organization provides them. Many organizations provide services which cannot be highly distinguished. Such organizations need to maintain and increase their quality of service. They use customer relationship management to acquire new customers and maintain the relationship with the old customers. It also helps to increase customer retention for more profits. The customer relationship management systems adopt analysis techniques for understanding customer behaviours. These analysis systems use machine-learning models to analyse the customers' personal and behavioural data to give organizations a competitive advantage by increasing customer retention rate. When an organization tries to maintain customer relationships, different phases come into action like they have to identify customers based on geography or economic value, retain the customer and attract more customers. A customer relationship management system was successfully built that can perform analysis on different forms of customer data for the betterment of businesses.

View PDFchevron_right
9 VI June 2021 https://doi.org/10.22214/ijraset.2021.35180 International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 9 Issue VI Jun 2021- Available at www.ijraset.com Customer Behavior Analysis using Machine Learning Mrs. T. L. Deepika Roy1, C. Anish2, Y. Chandra Shekar3, B. Pooja4 1 Asst.Professor, Department Of Computer Science and Engineering, CMRTC, Hyderabad, Telangana, India 2, 3, 4 Student, Department Of Computer Science and Engineering, CMRTC, Hyderabad, Telangana, India Abstract: RFM (Recency, Frequency, Monetary) investigation is a demonstrated showcasing model for conduct based client division. It groups clients dependent on their exchange history – how as of late, how frequently and what amount they buy.RFM helps partition clients into different classes or groups to distinguish clients who will react to advancements and how. This RFM examination depends on a blend of three boundaries. For instance, we can say that individuals who spend the most on items are our best clients. A large portion of us coincide and think about something very similar. In any case, Imagine a scenario in which they were bought just a single time. Or on the other hand an extremely quiet past? Consider the possibility that they are done utilizing our item. would they be able to in any case be viewed as your best clients? Most likely not. Making a decision about client esteem from only one perspective will give you a mistaken report of your client base and their lifetime. That is the reason, the RFM model joins three diverse clients ascribed to rank clients. In the event that they purchased in the recent past, they get higher focus. On the off chance that they purchase ordinarily, they get a higher score. What's more, on the off chance that they spend greater, they get more focus. Thus, we Combine these three scores to make the RFM score. At long last we can portion the client data set into various gatherings dependent on this RFM score. Keywords: Market Basket Analysis, Statistical Analysis, Clustering. I. INTRODUCTION Recently, to diminish human endeavors and time burned-through in checking the items and gathering the data about the items we have the retail examination. Retail investigation centers around giving bits of knowledge identified with deals, stock, clients, and other significant perspectives significant for dealers dynamic measures. The order incorporates a few granular fields to make a wide picture of a retail business wellbeing, and deals close by and large territories for development and support. Basically, retail investigation is utilized to help settle on better decisions, run organizations more proficiently, and convey improved client assistance investigation. The field of retail investigation goes past shallow information investigation, utilizing procedures like information mining and information disclosure to disinfect datasets to produce significant experiences that can be applied for the time being. Besides, organizations utilize these investigations to make better depictions of their objective socioeconomics. By harnessing deals information examination, retailers can recognize their ideal clients as indicated by diverse categories like age, inclinations, purchasing examples, area, and the sky's limit from there. II. LITERATURE SURVEY Smart marketers understand the importance of “know thy customer.” Instead of simply focusing on generating more clicks, marketers must follow the paradigm shift from increased CTRs (Click-Through Rates) to retention, loyalty, and building customer relationships. The concept of RFM was originally introduced by Bult and Wansbeek in 1995. It was used effectively by catalog marketers to minimize their printing and shipping costs while maximizing returns. An RFM analysis might not be the simplest way to segment your customers, but it doesn’t have to be too complicated. In this piece, we’ll go over how to create an RFM analysis, what an RFM score means, and what strategies to assign to those scores. RFM analysis is a marketing technique used to quantitatively rank and group customers based on the recency, frequency and monetary total of their recent transactions to identify the best customers and perform targeted marketing campaigns. The system assigns each customer numerical scores based on these factors to provide an objective analysis. RFM analysis is based on the marketing adage that "80% of your business comes from 20% of your customers." ©IJRASET: All Rights are Reserved 945 International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 9 Issue VI Jun 2021- Available at www.ijraset.com III. PROPOSED SYSTEM AND ARCHITECTURE RFM thinks about recency, recurrence and money related qualities for every client. Join them and afterward bunches them into distinctive item sections for simple review and crusade targeting. RFM investigation is super valuable in understanding the responsiveness of our clients and for division driven data set advertising. Advancing new items to steadfast clients is an extraordinary route for getting introductory footholds and criticism. You can contact your Loyal Clients even prior to building an item. They can give you extraordinary experiences into what to assemble and how to advance it. This gathering of individuals will likewise cheerfully allude your item to their circles of influence. FIG-1 System Architecture A framework design is the model that characterizes the construction, conduct and more perspectives on the framework. It contains framework parts and the framework created that will cooperate to carry out the general framework. In the Machine learning design , the crude information goes through numerous means prior to giving the yield. The entire information is screened and irregularities and indistinguishable information is eliminated from the informational index bringing about the yield with complete precision. IV. IMPLEMENTATION A. Data Cleaning The three separate datasets will be consolidated and examined. 1) EDA: Investigating the information and examining each segment and taking in experiences from fundamental information given utilizing different python perception libraries. 2) Null Worth and Outlier Treatment: This is a basic advance to perform during preprocessing the information, where the invalid qualities are checked and will be treated before any further assessment. And the anomalies will be treated considering their significance and the area understanding. 3) Feature Extraction: For the bunch examination ,utilizing entire information isn't significant and hence, separating not many highlights from the entire information and performing the assessment. 4) Cluster Investigation: Isolating entire information furthermore, shaping gatherings to draw out the top most items having extraordinary interest and deals. To check for low interest items furthermore, give important proposals. ©IJRASET: All Rights are Reserved 946 International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 9 Issue VI Jun 2021- Available at www.ijraset.com B. Information Cleaning As a business develops and develops,the size, number, designs, and sorts of its information resources change alongside it. Developments in finance frameworks, new organization equipment What's more, programming, arising production network innovations, and the like would all be able to make the need to relocate, consolidation, and join information from numerous sources."Dirty" information — information that contains redundancies, incorporates copy records, is missing data, or has been generally debased in the way toward being imported or combined — is one unavoidable outcome. Information change , which includes ―massaging information to make its fields and arrangements adjust to those of its objective, can likewise be the wellspring of hair pulling and restless evenings. C. Exploratory Data Analysis In measurements , exploratory information examination (EDA) is a way to deal with breaking down informational collections to sum up their primary attributes, regularly with visual strategies. A factual model can be utilized or not, yet basically EDA is for seeing what the information can advise us past the formal displaying or speculation. Exploratory information examination was advanced by John Tukey to urge analysts to investigate the information, and conceivably figure speculations that could prompt new information assortment and trials. EDA is unique in relation to starting information examination (IDA), which centers all the more barely around checking suppositions needed for model fitting and speculation testing, and taking care of missing qualities and making changes of factors as needeInsights about the purchase behavior. TABLE-1 Product Categories V. RESULTS AND ANALYSIS A. RFM Score Analysis Results The Result is represented in graphical format. Fig-2 RFM Analysis Graph ©IJRASET: All Rights are Reserved 947 International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429 Volume 9 Issue VI Jun 2021- Available at www.ijraset.com B. Clusters Formed After RFM Analysis Fig-3: Formed Clusters C. Graphical Representation Fig-4: Graphical representation of clusters VI. CONCLUSION With the ML Algorithms like grouping calculations and comparable item suggestions utilized, the arrangement was constructed easily giving advantages to the steadfast clients. These techniques were recognized easily utilizing the AI strategies in these applications. This methodology can be utilized in any retail market deals information where understanding client prerequisite also, building procedures that would increment deals to the organization and furthermore fulfill the clients that should purchase the items by giving offers, limits and integral or combos to the clients. VII. ACKNOWLEDGEMENT The authors would like to acknowledge the support of the Chairman, Director and Head of the Department, Department of Computer Science and Engineering, CMR Technical Campus, Medchal, Hyderabad, Telangana, for their encouragement to the authors. REFERENCES [1] RFM Analysis for Customer Segmentation | CleverTap [2] RFM Analysis For Successful Customer Segmentation - Putler [3] How to Use an RFM Analysis to Take Segmentation to the Next Level (omnisend.com) [4] What is RFM (recency, frequency, monetary) analysis and how does it work? (techtarget.com) ©IJRASET: All Rights are Reserved 948

References (3)

  1. RFM Analysis for Customer Segmentation | CleverTap
  2. RFM Analysis For Successful Customer Segmentation -Putler
  3. What is RFM (recency, frequency, monetary) analysis and how does it work? (techtarget.com)
About the author
Deepika Roy
Papers
6
Followers
1
View all papers from Deepika Royarrow_forward

Related papers

RFM Analysis to Understand Customer Patterns, Engagement and Retention in E-Commerce
K. N. Amarnath, Publishing India Group

International Journal of Business Analytics and Intelligence, 2023

Recency, Frequency and Monetary (RFM) analysis is a popular marketing technique that seeks to classify customers according to the frequency and value of their transactions. This article provides an introduction to the RFM analysis process and how it is used to gather crucial information for developing targeted marketing strategies. By using the RFM analysis on our catalogue and analysing sales data, we gain significant insights into the performance of particular products. According to the demand, the Recency dimension reveals which items are selling swiftly or slowly. Frequency dimension shows how well-liked particular goods are. The financial component, sheds information on viability and worth of any offering. RFM analysis gives marketers an opportunity to pinpoint various consumer categories with varied degree of loyalty, engagement and profitability. RFM analysis enables marketers to group customers into categories, like high-value, at-risk or dormant, using a combination of statistical methodologies and data mining. With a deep understanding of catalogue and sales data, we can make better choices on the products for customers. By emphasising top-selling items on the home-page, we may captivate customers and instantly pique their curiosity. In order to increase consumer involvement, we might simultaneously prepare to promote slower moving products with targeted incentives. The messaging, promotions and customer interactions can be modified for specific preferences of each group. High value customers can receive exclusive incentives to strengthen loyalty, and at risk customers receive re-engagement initiatives. By using RFM framework, businesses may develop targeted marketing strategies, enhance client interaction and boost marketing return-on-investment.

View PDFchevron_right
Understanding Customer Behaviour with Machine Learning
Rohan Bali

International Journal of Recent Technology and Engineering

The study of customer behavior both in online and offline purchases plays a very important role for the seller. The aim of this study is to identify customers on various parameters and thus re-define policies based on the behavior of customers. This paper works on churn analytics for retaining customers, a market-based analysis for identifying the support and confidence among products and a recommendation system built on the IBCF approach. Churn Analytics helps the seller to answer about whether the customers are leaving there products or services. The goal of every seller is to maintain a low churn rate and thus have large margins and bigger profits. Further, performing a marketbased analysis can be very fruitful for a supermart. This approach helps in organizing the items in a store in an efficient and scientific manner. This paper uses different machine learning algorithms techniques to conduct churn for the given data. It then calculates the accuracy and precision of each model ...

View PDFchevron_right
Customer Segmentation and Future Purchase Prediction using RFM measures
AKASH PATRA 21122005

International Conference on Advances in Computing, Communication Control and Networking (ICAC3N), 2023

Winning in the E-Commerce business race at a competitive age like this requires proper usage of Customer data. Using that database and grouping it in similar segments in terms of spending expenditure, observation time, sex, and location so that every customer falls in a segment of characteristics. This mechanism is called Customer Segmentation. In the modern era of highly compatible technological advancements, Machine Learning Algorithms are being vastly used to bring solutions to these difficult yet essential services. In the field of research methods like simple clustering based on purchase behavior, buyer targeting or automated customer promotion mechanism by dividing into two major categories, have been worked on. However, ensemble algorithms have come handy where different clustering algorithms are combined to deliver best segmentation. Lately combination techniques like clustering and classification mechanism have also delivered good results where, not only segmentation is done but also classification of existing and new customers are possible into the clusters. Depending on that an effective customer relationship management can really benefit the company to a huge extent. Unlike other studies where clustering was performed directly on RFM table, a different approach was taken in this study where, one dimensional clustering was done individually on Recency, Frequency, Monetary columns, then an overall score was calculated and customers were classified into three segments. However, for a new customer depending on his purchase behavior he/she also can be classified into any of the categories.

View PDFchevron_right
Retail Industry Analytics: Unraveling Consumer Behavior through RFM Segmentation and Machine Learning
Sydul Arefin

IEEE, 2024

This study explores the complex dynamics of consumer behavior in the UK retail industry using advanced analytics. We developed a preprocessed retail dataset including 495,478 retail customers. Here, we employ a combination of methods including a marketing analysis method known as RFM (Recency, Frequency, Monetary) segmentation, Box-Cox transformation, and temporal analysis to predict consumer behavior. Further, we use multiple supervised machine learning models (e.g. RandomForest, AdaBoost, ExtraTrees, LGBM, and XGBoost ) where XGBClassifier and ET have achieved the highest accuracy in forecasting customer lifetime value. We show that the performance metrics of ML models are impressive: an accuracy of 92.40%, precision of 92.27%, recall of 92.40%, F1 score of 92.28%, and an AUC score of 97.39. We also find that the results not only demonstrated the model's accuracy in forecasting clusters of customer lifetime value but also confirmed the robustness of ML methods. Key findings from RFM analysis indicate its unique merit in providing vital insights into clients and their behavior. This research establishes a new standard in retail analytics, offering a scalable and efficient methodology for future studies focused on utilizing data analytics to comprehend and predict customer behavior in different business entities.

View PDFchevron_right
IRJET- Customer Segmentation using RFM analysis
IRJET Journal

IRJET, 2021

Customer performs a transaction process on a daily basis. There are 56,799 transactions in the process, which generates a lot of data. The goal of this research is to execute consumer segmentation using a data mining approach based on the RFM (Recency, Frequency, and Monetary) model and clustering techniques. Based on the RFM model, a total of 56,799 transactions were performed. Based on the RFM model, 910 best Customers were identified from 56,799 transactions that were conducted. We also looked at clusters, and found that Cluster 2 had 16,335 customers, Cluster 3 had 11,342 customers, Cluster 4 had 341 customers, Cluster 5 had 294 customers, and Cluster 6 had 782 customers. The findings of this study can be used by a corporation to determine customer categories, and then the business determine how to sustain customer engagement.

View PDFchevron_right

Related topics

  • Computer Science
    • 580 California St., Suite 400
    San Francisco, CA, 94104
    © 2026 Academia. All rights reserved