Analyzing the Corporate Business Intelligence Impact: A Case Study in the Financial Sector

Abstract

Business intelligence is the process and methods that enable businesses to effectively analyze large amounts of data and transform it into meaningful information, helping to increase efficiency and productivity in businesses, thus enabling businesses to gain competitive advantage. In this context, business intelligence improves data management and decision-making processes and plays a critical role in strategic management. The main purpose of this study is to analyze the transition process of business intelligence solutions in financial institutions in detail, to increase efficiency in reporting processes, and to optimize decision-making processes. The study examines the ‘Cheque Report’, which reports the status of cheques in XY Financial Institution. Within the scope of the study, the transition process to business intelligence in the financial institution examined the ‘Cheque Report’ in three stages: in the first stage, reports were prepared manually; in the second stage, they were prepared with PL/SQL, and in the last stage, they were prepared with a business intelligence solution, and their outputs were compared. As a result, it was observed that with the use of business intelligence, fast and direct access to reports, data security, freedom from person dependency, and efficiency in internal information sharing are provided.

1. Introduction

In today’s increasingly competitive environment, institutions that effectively manage customer interactions are gaining strength in gaining a competitive advantage. Institutions dominate various data flows through management and analysis teams. It is of great importance for institutions to analyze existing data sources and transform these analyses into valuable information. The large amount of data presented to managers may lead to incorrect strategic decisions due to the insufficiently detailed analysis of these data. Therefore, institutions should have the ability to make future predictions. Institutions should be able to combine data from multiple sources into a single central source, analyze them in depth with analytical tools, establish predictive models for the future, and make decisions according to these models. Business intelligence plays an active role in this direction [1,2].

Business intelligence consists of tools and technologies used to make better business decisions. Business intelligence technologies include query, reporting, analysis tools, online analytical processing, decision support systems, dashboards, data mining, company performance management, company scorecard, and predictive analysis [3,4].

Business intelligence systems are information systems tools that are preferred by organizations and that obtain structured, semi-structured, or unstructured data and information, analyze them, and present them to the user. These tools enable effective decisions to be made by using operational data in organizations, making these data meaningful and presenting them to managers [5].

Business intelligence is a process and technology field that helps organizations analyze their data to obtain meaningful information and make better and more accurate decisions [6,7,8]. The software used in this field includes various tools to collect, store, analyze, and report data. Business intelligence helps organizations improve their performance by transforming data into meaningful pieces, understanding market trends, examining customer behavior, and optimizing business processes.

The main components needed in a business intelligence system are classified as OLAP, data mining, data warehouses, and ETL tools [9,10]. Online analytical processing (OLAP) cubes create data models to present business intelligence data in a meaningful way [10]. ETL (Extracting–Transforming–Loading) brings together data from different sources, transforms it when necessary, and loads it. Data warehouses are places where processed and organized data are stored [11,12]. These areas are optimized for fast querying and analysis. They are used to manage and store large data sets. Data mining uses data analysis and software techniques to find patterns, relationships, and structures within large data sets [13].

Williams and Williams (2010) defined business intelligence as business analysis, which directs decision-making processes and affects organizational mobility as the main purpose [14]. In the literature, business intelligence is considered by most researchers as an important component of decision-making processes. Al-Shubirii, in his study, investigated the impact of business intelligence on performance and presented an empirical case study [15]. Fernandez et al. (2016) presented a case study conducted with an SME in the northwestern part of Mexico [16]. They used IBM Cognos Analytics as a business intelligence tool in the study. As a result, an easy and fast data comparison was provided with the rack and stack approach. Owusu (2017) conducted a study on using a balanced scorecard in Ghana [17]. The study used SmartPLS as a business intelligence tool, and as a result, an indirect, positive, and significant growth trend was observed in learning and growth, internal business process, and customer performance with the adoption of BI systems. Abadi (2017) examined the implementation and impact of business intelligence tools, especially SAP Business Object and INFOR BI, in a trailer service company. called TIP Trailer Services [18]. The company offers trailer rental and sales, tanker rental and maintenance, tire management, damage repair, fleet management, technology and telematics, fleet optimization consulting, and trailer sales services. The research investigated operational reporting processes and the effectiveness of these business intelligence applications on decision-making and improving operational efficiency through a case study analysis using a qualitative methodology. Alzeaideen (2019) examined the critical role of credit risk management in the Jordanian banking sector and highlighted the need for business intelligence-based decision support tools to improve the accuracy and efficiency of credit approval processes [19]. The research addressed the limitations of methods that are often based on the subjective judgments of credit officers and traditional credit scoring models in credit evaluation processes. Richards et al. (2019) focused on the relationship between business intelligence effectiveness and corporate performance management practices [20]. As a research method, data were collected by distributing an online survey to senior managers in collaboration with industry partners PricewaterhouseCoopers and the Canadian Advanced Technology Association. An online survey consisting of 1300 senior managers was conducted, and 337 completed responses were obtained. The survey aimed to evaluate the effectiveness of business intelligence system implementation and management practices. In their study, Ahmad et al. presented business intelligence modeling by taking into account individual, technological, organizational, and environmental determinants in a systematic literature review of 84 articles published from 2011 to 2020 [21]. Tunowski (2020) aimed to determine if business intelligence (BI) systems contribute to the sustainability of commercial banks by affecting their financial health [22]. To evaluate this effect, a new comparative method was employed, analyzing financial condition indicators across three dimensions. The study found that BI systems positively influence various aspects of commercial banks’ financial health, specifically in terms of productivity, asset and liability quality, profitability, and debt. Akel et al. (2021) supported business intelligence applications with machine learning techniques and used them to visualize the obtained patterns to assist decision processes [23]. Zelenka and Podaras (2021) aimed to improve business intelligence processes by addressing the critical importance of data-driven decision-making processes in modern companies [24]. In their study, they focused on factors such as the effectiveness of business intelligence tools and data quality. They provided solutions specifically for improving data quality and aimed to strengthen the data understanding process with special information layers integrated into the existing business intelligence infrastructure with the proposed method. They also emphasized the importance of using information-based data to support collaborative decision-making processes in big data environments. In this study, conducted on a real example from the banking sector, they demonstrated the concrete benefits offered by the proposed information layers to businesses. Chen et al. (2021) examined the relationship between the development of business intelligence capabilities of businesses and firm performance [25]. The main purpose of their research was to experimentally verify the validity of the Sense–Transform–Drive conceptual business intelligence model developed based on dynamic capabilities and organizational evolutionary approaches. With this purpose, they determined the effects of business intelligence applications in the real economy. Niu et al. (2021) proposed Optimized Data Management using Big Data Analytics to increase smart corporate effectiveness and decision-making analysis in organizations [2]. In their study, a backtracking method was introduced in business intelligence and decision-making environments to increase plan failure and risk-taking ability. Shao et al. (2022) proposed an IoT-based Efficient Data Visualization Framework (IoT-EDVF) to enhance the risk of leakage, analyze multiple data sources, and strengthen data quality management for business intelligence in the field of corporate financial [26]. Alsibhawi et al. (2023) introduced a conceptual framework that outlines the key factors affecting the adoption of business intelligence systems within Libya’s SME sector [27].

Artificial intelligence has established itself as a crucial element in the decision-making framework of business intelligence. The integration of AI into BI systems has transformed how organizations gather, analyze, and understand data to make well-informed choices. Utilizing machine learning, natural language processing, and sophisticated analytics, the collaboration between AI and BI has enhanced decision-making, boosted efficiency, and strengthened competitive positioning [28,29]. Alghamdi and Al-Baity (2022) provided a comparison between traditional business intelligence (BI) and its augmented variant within the business analytics process [29]. Their findings indicated that augmented analytics improves analytical capabilities, shortens time frames, and facilitates data preparation, visualization, modeling, and insight generation. Siddiqui et.al. (2024) presented three models based on various feature set combinations and machine learning algorithms for predicting credit card customer churn in their study [30]. The results highlight the promise of machine learning in predicting customer attrition, allowing for the implementation of proactive retention strategies in the banking sector.

According to research conducted by Fortune Business Insights (2023) in the United States, the areas where business intelligence is most widely applied are banking, financial services, and insurance (BFSI), healthcare, information technology and telecommunications, retail consumer groups, and manufacturing. According to research, BFSI is expected to have the highest market growth during the forecast period [31]. Again, in the research conducted by Fortune Business Insights (2024), it was stated that the size of the global business intelligence market is expected to increase from USD 31.98 billion in 2024 to USD 63.76 billion in 2032 and has a compound annual growth rate of 9.0% during the forecast period (2024–2032) [31].

This study aims to examine how data-driven strategies and decision-making processes are shaped in the corporate world by focusing on the increasing importance of business intelligence today. Another focus of the study is to reveal concrete applications of the impact of business intelligence on corporate performance and how business processes are optimized. In this way, the contribution of business intelligence to corporate success will be understood more comprehensively and will enable businesses to understand how to best benefit from this technology. To the best of our knowledge, no detailed study has been found in the literature. Therefore, this study will contribute to eliminating this deficiency in the literature.

This study consists of five sections. After the introduction, the proposed research framework, methodology, and case study are examined. The results and discussion are presented in the Section 3, and the conclusions are given in Section 4. Finally, in Section 5, recommendations for future research are presented.

2. Proposed Research Framework, Methodology, and Case Study

2.1. Purpose and Importance of the Study

A cheque is a type of promissory note that is a negotiable instrument and contains a three-way exchange relationship and right to claim between the beneficiary, drawer (issuer), and interlocutor (bank).

The analysis of cheques used in financial institutions is an important topic in the financial sector and has been the subject of many studies. Although cheques are a widely used tool in payment transactions, monitoring and analyzing the payment status and number of cheques is critical to financial institutions’ liquidity management, risk management, and understanding of customer behavior. These analyses usually include determining the number and amount of cheque payment statuses. When examining the payment status of cheques, details such as the amount of cheques paid, the amount and reasons for rejected cheques, delays in the clearing process, and the dates of transactions are taken into consideration. In addition, customers’ cheque usage habits, the purposes for which cheques are used, and the payment performance of cheques are also analyzed.

There are many benefits to banks when they collect and analyze cheque-related data, including identifying customer risks, preventing fraud, improving liquidity management strategies, and improving service quality. These analyses can help banks improve their operational efficiency and provide better service to customers.

Within the scope of this research, it is aimed to increase the efficiency of the manually prepared reporting processes at XY Financial Institution and to encourage the use of business intelligence solutions. For this purpose, the study examines the implementation of a three-stage transition in the reporting process (preparing the reports manually in the first stage, preparing them using PL/SQL in the second stage, and preparing them with a business intelligence solution in the last stage). It aims to compare the outputs of the use of business intelligence with this three-stage transition. The application steps aimed at the study are presented in Figure 1.

2.2. Credit Registry Bureau (CRB) Cheque Report

It is a service that instantly queries cheques accepted for collateral or collection purposes. This service is provided with data, such as the cheque’s bank code, branch code, cheque account number, and cheque sequence number, and provides information about whether the cheque is in circulation. It is a report that aims to provide more comprehensive information about the validity and potential fraud of a cheque by verifying the identity of the cheque owner.

With the cheque query made via the CRB Cheque Report, a record is added to the relevant tables below after each query. Reporting is made based on the records added to the tables after each query process.

• CEKSOR Table: This is the table that holds the information about the queries that are recorded with a different ID as a result of each query made via the CRB Cheque Report screen.
• CEKICMAL Table: A SORGUNO is created in the CEKICMAL table for each ID field in the CEKSOR table. This table summarizes information about the number of cheques, payment status, and cheque status.
• CEKBANKA Table: A SORGUNO is created in the CEKBANKA table for each ID field in the CEKSOR table. This is the table that stores information about the bank name and bank code of the bank.
• CEKOZET Table: A SORGUNO is created in the CEKOZET table for each ID field in the CEKSOR table. This is the table that stores information about the maturity, amount, number, and status of the cheque.
• CEKKESIDECI Table: A SORGUNO is created in the CEKKESIDECI table for each ID field in the CEKSOR table. This is the table that stores information about the average cheque amount and the bank.
• CEKMUHABIR Table: A SORGUNO is created in the CEKKESIDECI table for each ID field in the CEKSOR table. In case of an error regarding the cheque, this is the table where the error code and error description information are kept.
• CEKSORSONUC Table: A SORGUNO is created in the CEKKESIDECI table for each ID field in the CEKSOR table. This is the table where detailed information such as the cheque reference number and transaction result is kept.
• CEKKESIDECITUTARI Table: A SORGUNO is created in the CEKKESIDECI table for each ID field in the CEKSOR table. This is the table where the year, type of cheque, and minimum and maximum amount of information regarding the cheque are kept.

The main table is the CEKSOR table. The ID correspondence formed by the CEKSOR table is created in the other tables as SORGUNO, and the matches between the tables are made through these fields.

The cheque tables in

Table 1. Number of columns and total number of records in cheque tables.

Table Names	Primary Key	Number of Columns	Number of Records in the Table
CEKSOR	ID	25	19,868,459
CEKICMAL	QUESTIONNO	26	738,159
CEKBANKA	QUESTIONNO	32	1,863,189
CEKOZET	QUESTIONNO	45	14,868,459
CEKKESIDECI	QUESTIONNO	15	2,963,256
CEKMUHABIR	QUESTIONNO	36	3,125,912
CEKSORSONUC	QUESTIONNO	25	7,074,994
CEKKESIDECITUTARI	QUESTIONNO	25	3,034,269

, which contain millions of records, demonstrate the functionality and value of big data analytics. These large data sets are too complex to be processed by traditional methods, and big data analytics analyzes these data to make strategic decisions, increase operational efficiency, and gain competitive advantage. Therefore, big data studies allow businesses to operate more effectively and efficiently.

2.3. Manual Preparation of Reports and Definition of the Problem

In the sales coordination department of XY Financial Institution, customer-based cheque statuses, the number of open cheques, cheque amounts, and the status of payments based on these are reported every week to be presented to management by querying 12 months before the current day. The report is prepared by a different employee in the unit every week. The preparation of the report takes until noon on Mondays every week, but sometimes extends to end of the day, depending on the abundance of data. In addition, a control period is also activated after the report is prepared.

The time study for the Manual Report is provided in Figure 2. While determining the report preparation time at XY Financial Institution, the report preparation process was measured with a stopwatch over 10 working days. The periods for the 10 days are shown by drawing a time study graph. According to the results obtained from the graphic trend, the average duration of the manually prepared report was determined to be 4 h.

The current situation was analyzed, and a workflow diagram was created and presented in Figure 3. The following problems were encountered during the manual preparation of cheque reports:

• Inefficiency and time loss of the manual process: Cheque reports are prepared manually on a weekly basis, and it takes an average of 4 h to prepare a report. In this process, operations such as manually combining and editing data cause a great loss of time. In addition, manual processes create personnel dependency, and the continuity and accuracy of the prepared reports are affected in case of personnel change.
• High error rates and lack of accuracy: Data entry errors made during manual reporting negatively affect the accuracy of reports. This situation is seen as a significant problem, especially in strategically important reports such as credit risk analyses. High errors slow down decision-making processes and damage corporate reliability.
• Impact on strategic decision-making processes: The fact that reports used by senior management are not based on detailed analysis causes deficiencies in strategic decision-making processes. In particular, the failure to report critical information such as cheque payment status, delays, and customer payment habits quickly and accurately negatively affects corporate performance.
• Operational risks coming with increasing data load: The increase in data over time makes it difficult to maintain manual processes. Working on large data sets takes more time and increases the risk of manual errors. This may lead to further growth in operational risks in the future.

As a result, it was found that the manual reporting process was a significant problem that negatively affected the efficiency of the operational processes at XY Financial Institution. Therefore, it is anticipated that the report should be prepared within a systematic framework rather than manually. In this context, it is aimed to combine data from multiple sources into a single data set by pulling the CRB Cheque Report on a customer basis querying the 12 months before the date the report was received, and reporting the cheque number, cheque amount, cheque drawer, tax number, cheque query date, service information, and cheque payment status of the customer with the most up-to-date query in the same line. In this way, it is aimed to facilitate the evaluation and action of the relevant unit with a simple report. For this purpose, ETL (Extract, Transform, Load) processes must be passed to combine and process the data effectively.

2.4. ETL Processes

For the collected data to be used in reporting processes, the data from the data sources must go through the extract, transform, and load phases. These extract, transform, and load phases refer to the ETL processes followed. The ETL transfer for the tables used in the CRB Cheque Report at XY Financial Institution occurs daily.

2.4.1. Extract Phase

• Cheque Inquiry Trigger: When a customer presents a cheque to XY Financial Institution or when a cheque needs to be processed, the institution automatically sends an inquiry about the cheque to the CRB or the relevant source. This inquiry includes information such as the cheque’s validity, the unclosed cheque’s status, and the cheque’s payment status.
• Retrieval of Data from Data Source: XY Financial Institution typically sends a query to the CRB platform to collect cheque information. This platform sends the cheque number, the date the cheque was issued, and other necessary information. The response from the CRB includes the current status of the cheque and the payment history.
• Data Format: Data are retrieved from the platform via the Rest API in JSON format. These formats allow for the data to be retrieved in a structured format. Cheque-related data typically include the following types of information:

  ○

  Cheque number

  ○

  Issuer/institution information

  ○

  Cheque amount

  ○

  Cheque status (paid, unpaid, bounced, etc.)

  ○

  Cheque date

  ○

  Cheque payment status (e.g., bank approved, voided, etc.)

2.4.2. Transform Phase

• Data Cleaning
  • Correction of Incomplete or Incorrect Data: Cheque query data received from the platform may be incomplete or incorrect. In this case, the accuracy of the data is ensured. For example, a cheque status may be “unpaid,” but the date information may be missing, in which case the missing date information can be completed.
  • Conversion of Data Formats: Data usually come to CRB. In cases where the data are in different formats (JSON, XML, CSV, etc.), all data are converted to a standard format. It is ensured that cheque amounts and dates are displayed consistently.
• Data Normalization
  • In cheque query data, it is necessary to take into account appropriate cultural formats to harmonize date formats. This ensures that users from different geographic regions see and understand the date data correctly. Correctly adapting date formats improves the user experience by eliminating data incompatibilities and ensuring system accuracy. It also ensures that operations performed with date information produce accurate and error-free results.

2.4.3. Load Phase

• Loading Data
  • Transfer to Data Warehouse: Collected and processed data are loaded into the bank’s data warehouse (DWH). These data are recorded in the CEKSOR, CEKICMAL, CEKBANKA, CEKOZET, CEKKESIDECI, CEKMUHABIR, CEKSORSONUC, and CEKKESIDECITUTARI tables in the DWH environment.
  • Real-Time Loading: Cheque query data are transferred at different times daily.
• Update and Maintenance
  • Data Updates: Cheque data are updated regularly. These updates are reflected in the DWH daily.
  • Data History and Monitoring: Cheque historical data are usually kept. In other words, the first query of a cheque and all subsequent query history are recorded. This plays an important role in the analyses performed in the future.

Figure 4 shows the data formation map. Here, the process starts with the issuance of a cheque. As a result of the entry of the issued cheque into the system by the financial personnel and the queries made from the CRB Cheque Screen, the data are instantly recorded in the relevant cheque tables. The data received instantly are included in the ETL processes, and the data warehouse (DWH) transfer, which is the source to be used in business intelligence solutions, is provided. Figure 5 shows the reporting process for the data transferred to the DWH.

2.5. Creating a Report Using PL/SQL Database

PL/SQL is a procedural language extension of the structured query language SQL developed for the Oracle database management system [32]. SQL is a standard language for data query and update operations in relational databases. The SQL-join operation is generally used to combine tables in querying the data we want via primary keys. There are multiple join options. However, since it is desirable to access all the data in the tables within the scope of this study, a restriction was not needed. Therefore, left-join matches were used.

In the database, the CEKSOR, CEKICMAL, CEKBANKA, CEKOZET, CEKKESIDECI, CEKMUHABIR, CEKSORSONUC, and CEKKESIDECITUTARI tables were connected by making a left join match over the ID and SORGUNO fields via the PL/SQL program, and technical analysis was conducted as a code for which approval was received for the relevant fields.

In Figure 6, a datamart was created by combining data from eight different tables under a single roof. Partition was added to the created datamart via the date field. This process provided ease of partitioning via the date field without changing the properties of the database and tables. In this way, it increases the processing speed by minimizing the CPU and memory consumption in these eight tables with large data.

2.5.1. Examining the Report Created Using the PL/SQL Database

The code created using the PL/SQL database is not a generic structure and is intended only to meet the request. In the following stages, when a change in the information in the report is requested, a coding study is needed. Systematically, the people working in the departments are not obliged to know PL/SQL from a technical perspective for reporting needs. The report created using the PL/SQL database does not provide a permanent solution in the long term in this respect. When performance was evaluated, the PL/SQL process plan (explain plan) was examined.

2.5.2. Explain Plan

The explain plan, a query plan that refers to a series of steps used to access data in SQL database management systems, is a special case of the relational model. A query plan is a roadmap that determines how a query will be processed by the database management system. This determines how the query will be optimized and which indexes or access paths will be used. In this study, the PL/SQL processing plan was examined in detail due to the size of the table and data. As seen in Figure 7, the running time lasted 13:31 min. This caused memory consumption warnings. It is expected that this period will be extended even further in the coming years with the increase in data in the table.

2.6. Creating the Report Using Business Intelligence Solutions

Business intelligence solutions were used because the report created using the PL/SQL database was not dynamic and had resource consumption warnings. Since there are multiple business intelligence solution integrated applications in the financial institution, while trying to select the business intelligence platform to be used, the expectations from the report are extracted in

Table 2. Report expectations.

Scope	Criterion
Data	Is there a need to obtain data from multiple data sources?
	Are different filters and variability needed?
Visuality	Will the user be presented with different visual options?
	Will reporting and analysis options change?
Sharing	Will it be used in collaboration with different units?
Applicability	Will it be adaptable to different demands?
Cost	High cost–performance ratio

. The expectations in Table 2 include items created through experiences gained from multiple requests made in different units.

After determining the information that should be included in the report, the opinions of the report-requesting units and project managers were taken and a score between 1 (minimum) and 5 (maximum) was made, and the most suitable business intelligence platform was selected among SAP BO, POWER BI, SAS BI, ORACLE BI, and QLIK. According to the election results, the platform that best fits the required criteria was determined as SAP BO with 32 points.

2.6.1. SAP Business Objects

SAP Business Objects (SAP BO) has different features and capabilities, helping businesses manage, analyze, and extract information from their data more effectively. SAP BO is a platform developed by SAP that offers a wide range of business intelligence and analytics solutions. The main purpose of SAP BO is to enable businesses to analyze data, create reports, and base their business decisions on better information.

SAP BO has the following capabilities:

• Data integration: Gathering data from different data sources and bringing it together.
• Reporting: Creating and sharing customizable reports.
• Analytics: Discovering trends, patterns, and statistics by analyzing data.
• Visualization: Making sense of data with charts, tables, and visual elements.
• Query: Accessing and analyzing data using queries.

In general, SAP Business Objects is a suite of powerful tools used to meet the data analysis and business intelligence needs of businesses. Businesses can use this platform to better understand their data, make business decisions, and gain a competitive advantage.

2.6.2. Web Intelligence Reporting and Universe Modeling

After selecting the business intelligence platform to be used, the Universe Design Tool was used to determine which tables should data be retrieved from and what kind of operations could be performed on them when reporting on the SAP BO platform. While requesters see the business layer on the front side of the Universe Design Tool, the master layer belongs to business intelligence users with administrator identity.

In the screen in Figure 8, the tables to be used in reporting on the master are added.

Figure 9 shows images of the tables added to the universe in the Information Design Tool (IDT) environment. These tables are presented simply on the tool without establishing relationships. Each table has a black title section at the top, and the column names in the table are indicated at the bottom. The following tables were selected from the existing tables in the database to be used for the CRB Cheque Report:

• CEKSOR
• CEKICMAL
• CEKBANKA
• CEKOZET
• CEKKESIDECI
• CEKMUHABIR
• CEKSORSONUC
• CEKKESIDECITUTARI

After the desired tables are brought to the front with the insertion process, the connections between the tables need to be created.

Among the tables in the database, the CEKSOR, CEKICMAL, CEKBANKA, CEKOZET, CEKKESIDECI, CEKMUHABIR, CEKSORSONUC, and CEKKESIDECITUTARI tables to be used in the CRB Cheque Report are selected. After the desired tables are brought to the front with the insert process, connections between the tables must be created. Figure 10 shows the universe table connection in the information design tool, and Figure 11 shows the business layer images of 1-N relationships in detail.

In the test conducted via Queries, a query was run by selecting the fields in the relevant cheque report folder, and it was observed that the fields matched and appeared in the report view. The fact that no error was received at this stage indicates that the relationships and matches established while adding columns were correct and that no errors would be received during the end user’s reporting. After determining which columns can be reported in the business layer and viewed by end users and that they work without errors via Queries, all of these should be saved for viewing on the Web Intelligence Rich Client. The Web Intelligence Rich Client is a platform used to visualize the columns selected for reporting in the Universe Design Tool.

The report provides detailed information on the cheque amount, total amount, number of cheques, service information, payment status, and maturity for each cheque. After the report is received, the summary of the report is displayed, and it is seen in Figure 12 that the report was received in 90 s.

3. Results

The work steps in each stage of the “three-stage transition process for reporting” aimed at in the study are shown in Figure 13.

When the reports obtained using the PL/SQL tool were examined, the following findings were obtained:

• The report is pulled using the PL/SQL tool via the datamart. However, the flexibility to select a column other than the columns written in the code was not provided.
• Since the report queries over the last 12 months, when data are to be pulled for older dates, it is necessary to raise a request for change to the business intelligence team each time for the code block that creates the datamart.
• Since the reporting process is via the database, the output is taken from the PL/SQL panel and presented by transferring it to Excel. Here, new work was needed to bring it back to the presentation format in Excel.
• Since the reporting tool is PL/SQL, only personnel with database authorization and usage authority could use it. This did not save the report from personal dependency.
• During the report extraction via PL/SQL, end users experienced bottlenecks in two stages while waiting. The first is the report query running time, and the second is the waiting period while transferring it to Excel and making final adjustments.
• Adding a graphic while getting the report from the panel screen on PL/SQL and the lack of visually satisfying reporting caused feedback from end users.
• Using the PL/SQL database, the CRB Cheque Report was reached in 13.5 min.

With integration into the business intelligence solution, the following findings were obtained:

• Without entering into a coding complexity such as PL/SQL, the report could be easily provided by the relevant teams by simply dragging and dropping the columns to the relevant fields.
• With the business intelligence solution, end users were able to change the report content in a short time by stretching it according to their needs.
• With the business intelligence solution, it was observed that visual colorings and graphics were added to the report within the same interface.
• With the business intelligence solution, the report could be recorded under a common record by authorizing the desired records, and its latest status could be presented to authorized persons.
• It was observed that the report obtained with the business intelligence solution can be obtained in 90 s and is in a format that can be used in presentation without the need for any intermediate program.

The time–cost graph Figure 14 clearly shows a comparison of reporting processes performed manually with those performed using PL/SQL and SAP BO. When the graph is examined, it is observed that each reporting tool shows significant differences in terms of weekly, monthly, and annual costs. In the manual reporting process, 240 min are spent weekly, 960 min monthly, and 11,520 min annually. This clearly shows that manual methods are inefficient, and reporting processes are carried out in a time-consuming manner. The high time cost of manual processes leads to a significant loss of productivity, especially in organizations working with large data sets. In the reporting process performed using PL/SQL, 13.31 min are spent weekly, 53.24 min monthly, and 638.88 min annually. This indicates a significant improvement compared to manual methods. PL/SQL makes the reporting process faster by automating database queries. However, a significant portion of time is still spent on manual interventions and database management. In the reporting process using SAP BO (Business Objects), 90 s are spent weekly, 6 min monthly, and 72 min annually. SAP BO minimizes manual interventions and significantly speeds up processes by providing automatic reporting and data analysis. This shows that SAP BO is the most effective solution, especially in terms of saving time. These findings show that business intelligence solutions provide great time savings and increase cost efficiency in the long term compared to manual and SQL tools.

Moreover, it is seen that the transition to business intelligence also provides significant improvements in error rates. In the manual reporting process, data collection and processing processes are largely based on the individual efforts of employees. In this process, cheque information is manually retrieved from the systems, and data from different sources are manually combined using Excel. This situation causes the processes to take a long time and results in a loss of time. The majority of errors that occur during manual reporting are errors made by employees in data entry or deficiencies in the file merging process. The error rate in the manual reporting process was calculated to be approximately 10%. With the introduction of PL/SQL, many operational errors in the manual process have been eliminated. Thanks to the processing of data directly with SQL-based queries and the provision of automation, the error rate has decreased to 3%. However, since PL/SQL-based systems generally require technical knowledge and do not offer a user-friendly interface, it has been difficult for employees to adapt, and the response time to customer requests has not improved to the desired level. With the implementation of the SAP BO business intelligence solution, the error rate has been reduced to 0.5%. Thanks to the visual reporting tools, user-friendly interface, and dynamic data integration offered by the system, the adaptation of employees to the reporting process has been facilitated, and reporting times have been significantly shortened.

A comparison table summarizing the key differences in efficiency and effectiveness for the targeted three-stage process is presented in

Table 3. Comparison table for the intended three-stage process.

Criteria	Manual Reporting	PL/SQL Reporting	SAP BO Reporting
Reporting Period	4 h	13 min	90 s
Error rate	10%	3%	0.5%
User Needs	No coding or SQL knowledge required	Requires coding and SQL knowledge	User-friendly, does not require coding or SQL knowledge
Working with Data Sources	Manual integration	Semi-automatic integration	Fully automatic integration
Dynamic Reporting	Not available	Not available	Available
Visualization	It is created via Excel	Not available	Automatic visualization and chart insertion
Specifications	Data is processed manually, time-consuming, and open to user error	Automation is provided with SQL scripts	It is the method with the fastest reporting time and lowest error rate.

.

4. Discussion

Adopting business intelligence (BI) tools plays a critical role in optimizing efficiency and reporting processes, especially in data-intensive industries such as the financial sector. On the other hand, various challenges are often encountered in the implementation of these systems. The recommendations for overcoming these challenges during the implementation of SAP BO in XY Financial Institution can be listed as follows:

(1) Overcoming Budget Constraints

• Evaluating Open-Source Solutions: The high costs of licensed tools such as SAP BO can be a major obstacle for small or medium-sized businesses. However, SAP BO was purchased in previous periods at XY Financial Institution, and the license cost was amortized in the short term considering the efficiency gains. However, companies that will use it for the first time can carry out the test stages with free or trial versions.
• Gradual Implementation: Instead of integrating BI tools into all systems at the same time, a gradual transition can control costs. First, a pilot application can be conducted by focusing on the most critical processes.

(2) Eliminating the Lack of Technical Expertise

• Organizing Training Programs: Tools such as SAP BO require technical knowledge. Therefore, regular training programs should be organized so that users can use BI tools more effectively. Practical training on SAP BO modules such as Information Design Tool and Web Intelligence will be effective.
• Consulting Services: External consulting services can be received to eliminate the lack of technical expertise during the transition process. This consultancy should cover not only technical integration but also the strategic use of BI tools.

(3) Overcoming User Resistance and Cultural Adaptation

• User-Friendly Tools: The user-friendly features of BI tools such as SAP BO should be emphasized and employees should be encouraged to use these tools. Visual reporting and drag-and-drop functions can be highlighted in particular.
• Creating Awareness: Awareness programs should be organized to explain the benefits of BI tools and how they facilitate processes. Employees’ understanding of the advantages that these tools provide them in terms of time and workload will accelerate adaptation.

(4) Integration of Business Processes

• Step-by-Step Implementation Plan: Integration of SAP BO into existing business processes should be done step by step with good planning in advance. Data integration processes in particular should be completed completely and correctly.
• Measuring the Process with KPIs: To evaluate the effectiveness of business intelligence systems, key performance indicators (KPIs) should be defined, and these indicators should be measured regularly.

(5) Transition Process Management

• Pilot Applications: Tools such as SAP BO should first be tested in a small department or a specific business process and then spread to other departments. This provides an opportunity to detect potential problems in advance.
• Creating Support Teams: An IT support team can be created to provide instant support to employees during the transition process.

5. Conclusions

Business intelligence plays an important role in the corporate world. It helps organizations determine more conscious and consistent strategies by enabling data-driven decisions [33]. It also enables strategic adjustments to be made towards targets by using key performance indicators in performance monitoring and evaluation processes [25]. In addition, business intelligence is effective in providing the opportunity to develop differentiation strategies in the market by analyzing the strengths and weaknesses of competitors in terms of gaining a competitive advantage [18]. It also provides opportunities to understand customer behavior and increase customer loyalty in customer relationship management [18,19]. Finally, it ensures data integrity and more reliable analysis by integrating information from different data sources [18,20,25].

As a result of this study,

• It was observed that corporate memory matured and its users increased.
• It was observed that errors in the reporting process decreased and data accuracy increased.
• Person dependency in the reporting process decreased significantly.
• It was observed that communication and information sharing within the financial department increased.
• It was determined that personal efficiency increased and cost advantage was provided.
• Thanks to the improvement in the preparation process, the efforts of the personnel were evaluated in different tasks.
• It was observed that with the widespread use of structured data, the trust of the users in business intelligence solutions increased, and thus the way for systematic improvements in time-consuming reports was opened.
• It was observed that the efficiency of business processes increased thanks to innovative reporting methods.

In future studies, middleware can be developed for cases where the same data are extracted at different times using business intelligence solutions and there is no change in the data structure. This software can automatically identify such situations. This approach can make it possible to remove unchanged data from reports, increase data accuracy, and make reporting processes more efficient. In addition, using middleware makes data management more efficient and can significantly contribute to the literature by improving the performance of business intelligence solutions.

That the experience of changing customer liquidity is based on the liquidity of loan/cheque repayments is missing in our model. After all, any delay in repayment can be an indication of the customer’s financial problems. To this end, the bank can change the loan/cheque repayment rules to help the customer. For this reason, we can mention in future studies that the “Customer Risk Analysis” module can be inserted into this framework to minimize the threats of returned cheques. Also, cheques may be used less for various reasons in the future. If this assumption is taken into account, loan repayments can be examined in future studies. Finally, the broader applicability of business intelligence to other sectors can be examined in future studies.