INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 251

CRM Boost: Predictive Sales Analytics for Enhancing Customer
Relations Using Regression Analysis

Rhonnel S. Paculanan, Geovanni P. Jomoc, Harold R. Lucero, Ronnie P. Gatdula, Ezekiel R. Borja, Laser Ryan V. Lleno

College of Computer Studies - Quezon City University

DOI: https://doi.org/10.51583/IJLTEMAS.2025.1409000035

Abstract: Predictive Sales Analytics for Enhancing Customer Relations Using Regression Analysis focuses on utilizing data-driven
techniques to inform and improve business decision-making. By applying regression analysis, the system can identify significant
relationships between customer behaviors and sales performance. This allows organizations to forecast future sales outcomes and
adjust strategies based on reliable predictions. The integration of predictive analytics into customer relationship management
(CRM) strengthens customer retention, loyalty, and satisfaction. It also helps managers allocate resources more effectively while
personalizing services to meet customer needs. The use of predictive models transforms raw customer and sales data into actionable
insights. Overall, this approach combines statistical accuracy with strategic value, enabling businesses to build stronger and more
sustainable customer relationships. The goal of this project is to design and evaluate a predictive sales analytics system that
leverages regression analysis to enhance customer relations. Specifically, the project aims to forecast sales outcomes, identify
customer trends, and provide actionable insights for decision-makers. By integrating predictive analytics into CRM practices, the
project seeks to improve customer retention, satisfaction, and loyalty. Furthermore, the system is expected to help businesses
allocate resources more effectively and achieve sustainable growth. A total of 150 respondents participated in the evaluation of the
system, consisting of 10 representatives from business organizations and 140 respondents from the academic community. The
diverse respondent composition allowed the system to be assessed from both practical and academic perspectives. The development
process was guided by the Agile methodology, ensuring iterative improvements and adaptability to user requirements. The
evaluation of CRMBoost was conducted using ISO quality standards, specifically focusing on functionality, interaction capability,
safety, security, and flexibility. Results showed that the system was functional in processing large amounts of customer data and
flexible enough to accommodate different organizational needs. Respondents confirmed that interaction capability was intuitive
and user-friendly, while safety and security measures were sufficient to protect customer information. The system outputs, which
included sales forecasts, customer trend reports, and predictive insights, were validated as reliable tools for both decision-making
and academic analysis. Findings revealed that regression analysis was effective in establishing relationships between independent
variables such as demographics, purchasing history, and seasonal demand, and dependent variables such as sales outcomes.
Business respondents highlighted the system’s practical usefulness in improving marketing strategies and customer engagement,
while the academic community emphasized its contribution to advancing knowledge in predictive analytics. Overall, the system
demonstrated its potential as both a technical solution and a strategic resource for organizations. In conclusion, CRMBoost bridges
the gap between theory and practice by combining predictive analytics with CRM strategies. It contributes significantly to
enhancing customer relationships while promoting sustainable business growth. The study recommends its adoption by
organizations and further development by researchers, with continuous alignment to ISO standards to ensure long-term
effectiveness and reliability.

Keywords: CRM, Predictive Analytics, Regression Analysis, Customer Relations, Sales Forecasting, ISO Standards, Agile
Methodology

I. Introduction

Predictive sales analytics has become a vital tool for businesses operating in today’s data-driven economy. Companies across
industries face the challenge of understanding customer behavior and anticipating market trends to remain competitive. Traditional
sales forecasting methods often fail to capture the complexity of modern consumer demands. To address this, organizations are
increasingly turning to advanced analytics techniques to generate accurate predictions. Predictive sales analytics provides a
framework to transform raw data into valuable insights that drive customer engagement and improve overall business performance.

Regression analysis has emerged as one of the most widely used techniques in predictive analytics. By examining relationships
between variables such as purchase history, demographics, and seasonal demand, regression models can generate reliable forecasts
of sales outcomes. Businesses that integrate regression analysis into their customer relationship management (CRM) systems are
able to identify key factors influencing customer behavior. This enables them to anticipate future purchases, segment customers
effectively, and personalize services. As a result, regression analysis contributes to stronger customer loyalty and sustainable sales
growth.

According to Byrapu, Surendranadha(2021) discussed how predictive models help businesses extract meaningful patterns from
customer data, which in turn improves decision-making. Regression-based predictive analytics was noted as an efficient approach
to forecast sales and optimize marketing strategies. Other scholars also emphasized that the integration of predictive analytics into
CRM allows organizations to proactively address customer needs. This shows that predictive sales analytics serves not only as a
technical tool but also as a strategic resource for customer engagement.

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 252

As presented by Teixeira, M. (2024) and Ahaggach, H., et al. (2024) demonstrate how regression models contribute to forecasting
accuracy in various business contexts. Their findings revealed that linear and multiple regression methods can establish clear
relationships between independent factors and sales performance. This provides organizations with the ability to predict future
demand more effectively. Additionally, regression models were found to guide decision-makers in adjusting promotional strategies
to match customer expectations.

Several related studies validate these theoretical insights by applying regression-based predictive analytics in real-world business
cases according to Ahmad, Jafar, Aljoumaa(2019) ) implemented regression models in a retail company and observed a 20%
improvement in customer retention rates. Similarly, In accordance to Schaeffer, Sanchez(2020) examined the telecommunications
industry and found that predictive models helped reduce customer churn by identifying at-risk clients in advance. From the study
of Mohaimin, Das, etal(2025) applied regression analysis to marketing campaigns and achieved more accurate targeting of
customers based on purchase behavior. These studies confirm that predictive sales analytics delivers measurable benefits when
applied to customer relationship management.

The integration of regression models into CRM has also transformed the way businesses approach personalization. Personalized
recommendations, promotions, and communication strategies are made possible by analyzing historical data. Customers are more
likely to remain loyal when they feel their preferences are recognized and valued. Predictive analytics allows businesses to deliver
tailored experiences, which not only enhance satisfaction but also improve sales performance. This level of personalization
represents a significant step forward in customer relationship management.

Predictive sales analytics contributes to better allocation of business resources. By forecasting demand and identifying key sales
drivers, organizations can optimize their inventory, marketing budgets, and workforce management. This ensures that resources are
directed toward areas with the greatest potential for impact. Businesses that adopt predictive analytics are therefore able to operate
more efficiently and effectively. This operational advantage further supports the argument for adopting regression-based predictive
sales analytics.

Predictive analytics has emerged as one of the most promising solutions for addressing customer demands and maintaining a
competitive edge. Regression analysis, in particular, provides the statistical foundation necessary for reliable predictions. By
combining accuracy with practical applicability, predictive sales analytics ensures businesses remain responsive to market
dynamics. This positions regression-based models as indispensable in modern business environments.

Scope

The goal of this project is to design and evaluate a predictive sales analytics system that leverages regression analysis to enhance
customer relations. Specifically, the project aims to forecast sales outcomes, identify customer trends, and provide actionable
insights for decision-makers. By integrating predictive analytics into CRM practices, the project seeks to improve customer
retention, satisfaction, and loyalty. Furthermore, the system is expected to help businesses allocate resources more effectively and
achieve sustainable growth.

Limitation

Despite its potential, the project has certain limitations that must be acknowledged. The accuracy of predictions largely depends on
the quality and quantity of available customer and sales data; incomplete or inconsistent datasets may affect results. Regression
analysis, while powerful, may not capture highly complex, nonlinear customer behaviors compared to advanced machine learning
algorithms. The system is limited to analyzing structured data such as purchase history and demographics, excluding unstructured
data sources like social media sentiment or customer feedback. Another limitation of the system is it is limited by its reliance on
historical data, which may not capture sudden market changes or shifts in customer preferences. Its effectiveness may also be
reduced in industries with rapid trends, as regression models can oversimplify complex relationships and ignore qualitative factors
like customer emotions. Adoption challenges arise due to the need for technical expertise, infrastructure, and integration with
existing CRM platforms, which may be difficult for smaller organizations. Data privacy and security risks further constrain the
system, alongside the possibility of biased predictions from flawed historical datasets. Ultimately, the system’s success depends
heavily on proper training and stakeholder engagement to maximize its potential benefits.

Theoretical Framework

This study is anchored on theories that explain how predictive analytics and regression models contribute to effective decision-
making in business contexts. The Customer Relationship Management (CRM) Theory emphasizes the importance of systematic,
data-driven strategies in fostering long-term customer loyalty and satisfaction, supporting the integration of predictive models to
anticipate customer behavior. The Predictive Analytics Theory posits that historical data can be examined to forecast future
outcomes, thereby providing businesses with actionable insights through meaningful patterns and trends. Meanwhile, the
Regression Analysis Model serves as the statistical foundation, establishing relationships between dependent variables such as sales
outcomes and independent variables such as demographics, purchase history, and market conditions. Together, these theories justify
the use of regression-based predictive analytics to enhance CRM by generating accurate forecasts and improving decision-making.

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 253

Conceptual Framework

The conceptual framework of Predictive Sales Analytics for Enhancing Customer Relations Using Regression Analysis
demonstrates how customer and sales data are processed through regression models to generate predictive insights. These insights
guide decision-makers in improving CRM practices, leading to enhanced customer satisfaction, retention, and sustainable business
growth.

Figure 1: Conceptual framework of CRM Boost: Predictive Sales Analytics for Enhancing Customer Relations Using Regression

Analysis

 Input: Customer and sales data such as purchase history, demographic profiles, seasonal demand, and past sales
performance serve as the primary inputs.

 Process: These data sets are analyzed using regression algorithms to identify significant variables, establish patterns, and
forecast future sales outcomes. The system processes data into predictive models that reveal customer trends and behavior.

 Output: The predictive results provide actionable insights for decision-makers, including sales forecasts, customer trend
analysis, and suggested marketing or resource allocation strategies.

 Outcome: Enhanced CRM outcomes such as improved customer satisfaction, increased retention, personalized services,
and sustainable business growth are achieved.

This framework positions predictive sales analytics as both a decision-support tool and a customer engagement enhancer. It
demonstrates how regression analysis bridges raw data and practical business strategies, aligning the project’s goals with
measurable outcomes.

Significance of The Study

This study is significant because it bridges the gap between advanced data analytics and customer relationship management (CRM)
practices. By applying regression analysis to sales and customer data, businesses can gain accurate forecasts of sales outcomes and
a deeper understanding of customer trends. These insights are crucial for designing strategies that enhance customer satisfaction,
loyalty, and retention.

For business organizations, the system provides a decision-support tool that helps allocate resources effectively, optimize marketing
campaigns, and achieve sustainable growth. For customers, the research ensures more personalized services, improved product
recommendations, and timely responses to their needs, resulting in stronger engagement and trust.

For the academic community, this research adds to the growing body of knowledge on predictive analytics and its applications in
business intelligence and CRM. It also provides a framework for future studies that wish to explore the use of regression and other
machine learning models in sales forecasting and customer relations. For IT practitioners and developers, the study serves as a
practical reference in designing analytics-driven CRM tools that integrate predictive modeling with real-world business
requirements.

Hypothesis of The Study

 Null Hypothesis (H₀): The integration of predictive sales analytics using regression analysis does not significantly enhance
customer relations in terms of satisfaction, retention, and loyalty.

 Alternative Hypothesis (H₁): The integration of predictive sales analytics using regression analysis significantly enhances
customer relations by improving satisfaction, retention, and loyalty.

Review of Related Literature

 According to Mirzaei & Iyer (2014) review and classify applications of predictive analytics in CRM and find that
regression methods help analyze customer behavior and support customer acquisition and retention decisions.

 As stated by Rathi, M., Vinu V. Das, and others (2010) demonstrate how regression modeling techniques applied to CRM
data can predict customer responses and maintain customer loyalty based on attributes like age, income, and gender.

 As emphasized by Umair Ejaz & Mathew Gimah (2024) show that predictive analytics, including logistic regression and
decision trees, significantly improve customer retention by identifying behavioral indicators of churn.

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 254

 As highlighted by the study “Big Data Analytics And Predictive Analysis In Enhancing Customer Relationship
Management (CRM)” by Khan, Akter, & Islam (2024) concludes that predictive modeling and real-time data processing
enable CRM systems to anticipate customer behavior and enhance satisfaction and retention.

 As pointed out by in “Application of data mining techniques in customer relationship management: A literature review
and classification” (2009), the authors report that regression and forecasting are among key data mining functions used in
CRM for customer retention strategies.

 According to the study of the paper “Enterprise Analytics using Graph Database and Graph-based Deep Learning” (Henna
et al., 2021) applies graph-based predictive models that outperform traditional regression models in CRM settings by
incorporating network features.

 As identified by Posedel Šimović, Horvatic, & Sun (2021) develop a mixed-penalty logistic regression model to classify
customers’ online engagement and improve churn prediction in CRM datasets.

 In the article “Artificial Intelligence and Machine Learning in CRM: Leveraging Data for Predictive Analytics” (Reddy et
al., 2021) shows how predictive analytics tools like regression can enable accurate forecasting, lead scoring, and
personalization in CRM.

Synthesis

The reviewed literature consistently highlights that predictive analytics plays a vital role in improving customer relationship
management by forecasting behaviors and identifying patterns. Several studies emphasized that regression analysis, whether linear
or logistic, provides accurate forecasting of sales outcomes and customer retention trends. Scholars agree that integrating predictive
analytics into CRM systems strengthens decision-making by offering actionable insights for both businesses and researchers. While
business organizations view predictive models as tools for competitive advantage and customer loyalty, the academic community
values them for advancing theoretical frameworks in data-driven management.

II. Methodology of The Study

Applied Technology Research was used to focus on developing a predictive sales analytics system that integrates regression
analysis into CRM practices for practical business use. Applied research emphasizes solving real-world problems, and in this case,
it addresses the need for accurate sales forecasting and improved customer relations. As demonstrated by Posedel Šimović,
Horvatic, & Sun (2021), predictive models using logistic regression with mixed penalties can classify different customer
engagement behaviors and improve churn prediction performance, supporting the bridge between statistical methods and practical
CRM decision-making.

Agile methodologies provide a flexible framework for developing CRMBoost by enabling iterative testing and continuous
refinement of regression models. This approach ensures that predictive sales analytics remains aligned with shifting customer data
and organizational needs, making the system more reliable and responsive (Beck et al., 2001). Furthermore, Agile fosters strong
collaboration between developers and business stakeholders, which enhances the system’s ability to deliver actionable insights that
improve customer relations and decision-making (Beck et al., 2001).

Figure 2: Agile Software Development Methodology

Through this approach, the system can be continuously refined to improve forecasting accuracy and strengthen customer
relationship management outcomes. Agile Software allows continuous improvement through iterative cycles. The process begins
with planning and requirement analysis, where system objectives and data needs are defined. This is followed by data collection
and preparation, ensuring that customer and sales records are accurate for regression analysis. In the development and testing stages,
regression models are built, integrated into the CRM system, and evaluated for accuracy and usability. Finally, the system goes
through deployment and maintenance, where it is implemented in a business setting and continuously refined to enhance customer
satisfaction and decision-making.

Respondents of This Study:

Composition of 150 individuals from two key groups. Out of this number, 10 respondents are from business organizations,
representing managers and decision-makers who utilize CRM and sales forecasting in their operations. The remaining 140
respondents are from the academic community, consisting of faculty members, researchers, and students with expertise or

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 255

background in information systems, business analytics, and customer relationship management. This composition ensures that the
study captures both practical industry insights and academic perspectives. Such a balance allows for a more comprehensive
evaluation of the predictive sales analytics system, addressing its relevance in both professional and scholarly contexts.

Development and Testing and The Iso 25010 Standard

The development of the system utilized modern web technologies to ensure efficiency and scalability. The front-end of the web
application was built using HTML5, CSS3, JavaScript, Alpine.js, Chart.js, and jQuery/AJAX to create an interactive and visually
appealing user interface. These technologies allowed for dynamic data visualization and seamless user interactions. For the back-
end, the system was developed using Python 3.9+ and FastAPI, with Uvicorn as the ASGI server for high-performance request
handling. SQLAlchemy and Pydantic were used to handle the application's data layer, ensuring reliable data modeling and
validation. Scikit-learn was integrated to implement regression analysis algorithms for predictive sales analytics, providing insights
based on historical sales data. The database system was built on PostgreSQL, chosen for its scalability, reliability, and efficient
handling of complex queries.

The system development and testing processes followed the ISO 25010 quality model to evaluate and maintain software quality.
The ISO 25010 standard was used to assess the system across multiple development and testing stages, ensuring optimal
performance and reliability. This elaborates the scope of development and testing under five (5) classifications:

 Functional Suitability: The degree to which a system, product, or component provides functions that meet stated and
implied needs under specified conditions.

 Interaction Capability: The degree to which a system, product, or component enables effective user interaction and
engagement.

 Safety: The degree to which a system, product, or component operates without causing unacceptable risks or harm.

 Security:The degree to which a product or system protects information and data so that users, other products or systems
have the degree of data access appropriate to their types and levels of authorization

 Flexibility: The degree to which a system, product, or component adapts to changes in environment and requirements.

Data Analysis Plan

The primary research instrument was a survey questionnaire. The proponents used weighted mean as a statistical method to evaluate
the system in accordance with the ISO 25010 criteria. Functional Suitability, Interaction Capability, Safety, Security, Flexibility
were used as criteria to evaluate the system. Frequency and Percentage Distribution was used to itemized breakdown of answers
to each question depending on the study’s categories or the set of responses defined in the survey, while the Weighted Mean is to
measure effectiveness, this rating scale was created utilizing a five-point Likert scale approach, with responses choosing one of the
five potential reactions from the choices for each question and the Likert Scale used for the interpretation on the evaluation based
on ISO 25010 criteria.

The System

The system output consists of comprehensive sales forecasts that help organizations anticipate future revenue trends. It also provides
detailed customer trend analysis, highlighting purchasing patterns, demographic behaviors, and seasonal demand. Managers receive
predictive reports and visual dashboards that support data-driven decision-making. Additionally, the system generates actionable
recommendations for resource allocation and marketing strategies. Overall, these outputs enhance customer relationship
management by improving satisfaction, retention, and loyalty.

Figure 3 Admin Dashboard Overview

This presents the Admin Dashboard Overview. This dashboard shows important information like the number of total users, active
users, and verified users. It also displays login activities, success rates, failed login attempts, and alerts for suspicious IP addresses.
The overview helps administrators quickly understand the system’s usage and security status, making it easier to manage users and
maintain platform safety.

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 256

Figure 4 User Management Interface

This shows the User Management Interface inside the Admin Panel. Here, administrators can see a list of all users along with their
names, emails, roles, and account status. Admins can easily edit user details, delete accounts, or filter users based on different
criteria. This makes managing the user database simple and organized.

Figure 5 Audit Logs page

This shows the Audit Logs page inside the Admin Panel. This page records all important user activities, such as when accounts are
activated or when users verify their emails. It includes details like the date of the action, the user’s email, what action was done,
and who performed it. There are also filter options at the top to help admins find specific activities quickly, making it easier to track
system events and ensure security.

Figure 6 Overview page of the Sales Dashboard.

This shows the Overview page of the Sales Dashboard. It gives a summary of sales activities, showing sections for Active
Opportunities, Total Pipeline Value, and Win Rate (although no data is entered yet). Below these metrics, there is a bar graph that
shows monthly sales results. This overview helps users quickly check sales performance and spot trends.

Figure 7 Forecast dashboard

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 257

This shows the Forecast section inside the Analyst Dashboard. It includes a line chart that predicts how sales numbers will change
over time. This forecast helps users plan by showing expected sales growth or decline based on current data and trends.

Figure 8 Analysis dashboard

This shows the Analysis section of the Analyst Dashboard. It features a bar graph that shows predicted growth rates for different
regions. This helps users see which areas are expected to perform better and supports better decision-making by providing insights
based on past data and predictions

Assessment: Summary of The Respondents on The System

The data shows that the researchers had set five (5) criteria for the respondents to assess the system: Functionality, Interaction
Capability, Safety, Security, and Flexibility. Based on the calculated results, the overall average weighted mean is 4.68.

Table 1. Assessment Summary of the respondents on the System

Criteria Weighted Mean Verbal Interpretation Rank

Functionality 4.73 Strongly Agree 1

Interaction Capability 4.66 Strongly Agree 1

Safety 4.66 Strongly Agree 1

Security 4.72 Strongly Agree 1

Flexibility 4.64 Strongly Agree 1

Average Weighted Mean 4.68 Strongly Agree 1

The table shows “Strongly Agree” as an overall interpretation. It clearly indicates that the proposed system successfully meets all
the specified criteria. It demonstrates that the system is functional, interactive, safe, secure, and flexible.

Ethical Considerations.

The study ensures that data collected from both business organizations and the academic community are handled with
confidentiality and integrity. Respondents’ information is protected, and no personally identifiable details are disclosed without
consent. The research adheres to principles of voluntary participation, where respondents are free to withdraw at any time without
consequences. Data security measures are strictly observed to prevent unauthorized access and misuse of information. Finally, all
findings are reported with honesty and accuracy, avoiding any form of manipulation or bias to maintain the credibility of the study.

Summary

The study aimed to develop and evaluate a predictive sales analytics system using regression analysis to enhance customer
relationship management. A total of 150 respondents participated, with 10 from business organizations and 140 from the academic
community, ensuring a mix of practical and academic insights. The system design followed the Agile methodology, allowing
iterative improvements throughout the development process. The evaluation of the system was guided by the ISO quality
framework, specifically assessing functionality, interaction capability, safety, security, and flexibility. Respondents confirmed that
the system met functional requirements and was capable of handling customer and sales data effectively. The system outputs
included sales forecasts, customer trend analysis, and predictive reports that supported better decision-making. Findings revealed
that regression analysis provided accurate forecasting and meaningful insights into customer behavior. The academic community
emphasized its contribution to knowledge, while businesses validated its real-world application. Overall, the study bridged
theoretical understanding with practical implementation of predictive analytics in CRM.

INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)

ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue IX, September 2025

www.ijltemas.in Page 258

III. Conclusion

Based on the findings, it can be concluded that predictive sales analytics using regression analysis is highly effective in enhancing
CRM practices. The system successfully forecasted sales trends, identified customer behaviors, and provided reliable insights for
resource allocation. By analyzing customer demographics, purchase history, and seasonal demand, the regression models
demonstrated accuracy and relevance. The involvement of both business practitioners and academic respondents validated the
system from different perspectives. Business representatives confirmed the tool’s practicality, while academic respondents
supported its theoretical soundness. The use of Agile methodology ensured flexibility in system development and allowed
refinement throughout the process. The system outputs, such as sales forecasts and customer trend reports, proved beneficial to
decision-makers in improving strategies. This highlights that predictive analytics is not just a technical feature but also a strategic
resource for competitive advantage. The study affirmed that CRM systems integrated with predictive analytics can strengthen
customer loyalty and satisfaction. Therefore, the system provides significant contributions to both academic research and industry
applications.

IV. Recommendation

It is recommended that businesses and academic institutions continue to adopt predictive analytics systems for CRM enhancement.
Organizations should invest in accurate data collection and management to maximize the functionality and flexibility of the system.
Respondents suggested that ISO standards—particularly interaction capability, safety, and security—must remain a priority to
ensure the ethical and reliable use of customer data. Future researchers may expand the study by comparing regression with other
advanced predictive algorithms. It is also recommended that the system outputs, such as sales forecasts and customer trend analyses,
be enhanced through real-time dashboards for easier accessibility. Agile methodology is further recommended for ongoing system
development to ensure adaptability in fast-changing business environments. Training programs for employees and academic users
should be introduced to maximize the potential of predictive tools. Additionally, collaboration between business organizations and
academic institutions is encouraged to refine the system and expand its applications. Finally, continuous evaluation against ISO
standards should be conducted to guarantee long-term sustainability, effectiveness, and customer satisfaction.

References

1. Abdelrahim Kasem Ahmad, Assef Jafar & Kadan Aljoumaa,(2019), Customer churn prediction in telecom using machine
learning in big data platform, Journal of Big Data (2019).

2. Ahaggach, H., et al. (2024). Systematic Mapping Study of Sales Forecasting. MDPI—offers a review of forecasting
techniques (including regression) and their evolution.

3. Beck, K., Beedle, M., van Bennekum, A., Cockburn, A., Cunningham, W., Fowler, M., Grenning, J., Highsmith, J., Hunt,
A., Jeffries, R., Kern, J., Marick, B., Martin, R. C., Mellor, S., Schwaber, K., Sutherland, J., & Thomas, D. (2001).
Manifesto for Agile Software Development. Agile Alliance. http://agilemanifesto.org/

4. Byrapu, Surendranadha Reddy Sr. Analyst (2021) Predictive Analytics in Customer Relationship Management: Utilizing
Big Data and AI to Drive Personalized Marketing Strategies Australian Journal of Machine Learning Research &
Applications

5. E.W.T. Ngai,Li Xiu, D.C.K. Chau (2029), Application of data mining techniques in customer relationship management:
A literature review and classification, Expert Systems with Applications

6. Khan, Akter, & Islam (2024), Big Data Analytics And Predictive Analysis In Enhancing Customer Relationship
Management (CRM), Journal of Machine Learning, Data Engineering and Data Science

7. Mirzaei & Iyer (2014), Application Of Predictive Analytics In Customer Relationship Management: A Literature Review
And Classification, 2014 Proceedings. 23.

8. Mohaimin, Das, Akter, Anonna, Hasanuzzaman, Chowdhury & Alam (2025) Predictive Analytics for Telecom Customer
Churn: Enhancing Retention Strategies in the US Market, Journal of Computer Science and Technology Studies

9. Petra Posedel Šimović, Davor Horvatic, Edward W. Sun,(2021) Classifying variety of customer's online engagement for
churn prediction with mixed-penalty logistic regression, arvix publication

10. Rathi, M., Vinu V. Das, and others (2010), Regression Modeling Technique on Data Mining for Prediction of CRM,
International Conference on Advances in Information and Communication Technologies

11. Ravi Teja Potla, Vamsi Krishna Pottla(2024), Artificial Intelligence and Machine Learning in CRM: Leveraging Data for
Predictive Analytics, Journal artificial Intelligence research

12. Satu Elisa Schaeffer, Sara Veronica Rodriguez Sanchez(2020), Forecasting client retention — A machine-learning
approach, Journal of Retailing and Consumer Services

13. Shagufta Henna, Shyam Krishnan Kalliadan, Enterprise Analytics using Graph Database and Graph-based Deep Learning”
(Henna et al., 2021), arvix publication

14. Teixeira, M. (2024). A Comparative Study Using ARIMA and Deep Neural Networks for Retail Forecasting — discusses
incorporating promotional data to improve forecast accuracy. MDPI Proceedings.

15. Umair Ejaz & Mathew Gimah (2024), Predictive Analytics for Customer Retention, researchgate