Starting out as a data analyst can feel overwhelming, especially when theory meets reality. One of the best ways I’ve found to grow is by working with real data sets-no matter how imperfect or small they may seem.

In this article, I’ll walk you through my first data analysis project, where I analyzed a Nigerian customer transaction and loan repayment data and built an interactive Excel dashboard. This project helped me understand how raw data can be transformed into meaningful insights that support business decision-making.

INTRODUCTION

Understanding customer transactions and loan repayment patterns is crucial for financial institutions in Nigeria. These insights help organizations manage credit risk, improve service delivery and enhance customer experience.

By analyzing transaction trends and repayment behavior, financial institution can better understand:

· Customer preferences

· Creditworthiness

· Levels of financial engagement

This project focuses on uncovering those insights using Excel as a beginner data analyst.

About the Dataset

The data was imported into Microsoft Excel and contains 301 rows and 13 columns. It includes the following fields:

· Customer ID

· Transaction Date

· Day, Month, Month Name, Year

· Transaction Type: Deposit, Bill Payment, Loan Repayment, Transfer, Withdrawal

· Channel: USSD, Web, Agent, Mobile App

· Amount (NGN)

· Customer Age

· Age Group

· Region

· Loan Repayment status (On-Time, Late, Not Applicable)

Key Analyses Conducted

Using this dataset, I carried out the following analyses:

1. Transaction type analysis

2. Channel usage analysis

3. Monthly transaction trend

4. Transaction by state

5. Loan repayment status analysis

6. Age group loan analysis

These analyses were designed to answer real business questions and provide actionable insights.

Data Cleaning and Preparation

Data cleaning was one of the most important and challenging parts of this project. Some of the steps I took include:

Formatting the transaction date into Day, Month, Month Name and Year.

Cleaning the Amount (NGN) column, which originally contained errors such as?8,033. I converted these into numeric value using the formula: VALUE(RIGHT(i2, LEN(i2)-1)).

Replacing blank entries in the loan repayment status column with N\A which means NOT APPLICABLE.

Verifying consistency and accuracy before visualization.

Only after cleaning the data, I proceed to analysis and dashboard creation.

Three important KPIs was developed to answer the questions:

What are the Total Transactions?

What are the Total Number of Customers?

What is the Average Transaction’s Amount?

1. Transaction Type analysis (Bar Chart)

2. Analysis of channel used (Bar Chart)

3. Monthly transactions trend (Line Chart)

4. Transaction by state analysis (Column Chart)

5. Loan Repayment status Analysis (Doughnut Chart)

6. Age group loan analysis (Column Chart)

Dashboard Overview

· Seasonal Trend: Transaction volume peak early in the year and experiences a sharp mid-year decline, followed by mild, uneven recovery toward year-end

· Loan Repayment Behavior: Repayment performance is almost evenly split among On-time (33.81%), Late (35.40%) and N/A (30.79%), indicating mixed repayment culture.

· Top Transaction Types: Deposits lead with 81 counts, followed by withdrawal (72) and Transfers (58)

· Regional performance: Abuja and Ibadan show the highest transaction totals, while Port Harcourt records the lowest.

· Channel Usage: Mobile App and Agent channels dominate customer engagement, while Web and USSD are moderately used.

· Loan Distribution: Adults take significantly higher loan amounts compared to youth.

This makes it easy for stakeholders to understand at a glance.

Challenges

Throughout the project, I encountered several challenges that helped me grow as a beginner data analyst:

1. Data Cleaning Difficulties
   Some entries — especially the transaction amount column containing symbols and errors — required additional cleaning steps. Understanding how to extract numeric values using formulas such as VALUE(RIGHT(i2, LEN(i2)-1)) took time but improved the accuracy of the dataset.
2. Handling Missing and Inconsistent Values
   The Loan Repayment Status column contained blank spaces that disrupted analysis. Identifying and replacing missing values with “N/A” was necessary to avoid errors during pivot table creation.
3. Choosing the Right Visuals for Each Metric
   Selecting the right chart types for KPIs and analysis — especially deciding when to use bar charts, line charts, doughnut charts, and columns — required trial and error to achieve clarity and accuracy.

4. Dashboard Layout and Formatting
Combining multiple pivot charts, slicers, and KPIs in one dashboard without overcrowding the view was challenging. Ensuring that everything remained interactive and easy to interpret demanded careful spacing and formatting

Insights

The analysis revealed several important patterns about customer transactions and loan repayment behavior:

1. Customer Transaction Behavior
   Deposits and withdrawals emerged as the most common transaction types, suggesting customers primarily use the platform for money movement and daily financial activities.
2. Preferred Transaction Channels
   The Mobile App was the most used channel, closely followed by Agent-assisted transactions. This shows that customers prefer convenience, accessibility, and human guidance over Web or USSD channels.
3. Seasonal Transaction Trends
   Transactions peak early in the year, decline sharply mid-year and show only mild recovery toward year-end.
4. State-Level Activity Differences
   Highly urbanized states like Lagos, Abuja, Ibadan, and Kano recorded the highest transaction frequencies, indicating higher digital finance adoption in densely populated areas.
5. Loan Repayment Behavior
   Most customers repaid loans on time, but among active loan users, late repayment was noticeably higher than expected — an indicator of moderate loan discipline.
6. Age-Based Loan Engagement
   Adults performed the highest number of transactions and showed stronger loan repayment engagement. Youths had fewer loan activities, suggesting lower borrowing behavior or financial participation.

Recommendations

Based on the insights generated, the following recommendations can support better decision-making:

1. Enhance Mobile App Features

Since most customers prefer using the mobile app, improving its user experience, speed, and security will further strengthen customer engagement.

2. Leverage Agent Network for Financial Inclusion

Agent transactions rank second in usage. Expanding agent presence in less active states can increase customer reach and support users who need physical assistance.

3. Promote Digital Awareness in Low-Activity States
Targeted campaigns and simplified onboarding can increase adoption in rural or low-transaction states.

4. Improve Loan Monitoring and Customer Support
With late repayment relatively high among active borrowers, offering reminders, flexible repayment plans, and financial literacy support can reduce default risk.

1. Introduce Youth-Focused Financial Products
   Low loan engagement among youth suggests a gap. Creating student-friendly or youth-oriented credit products may boost adoption.
2. Use Historical Trends to Plan Peak-Season Resources
   Since transactions spike early in the year, institutions should strengthen system infrastructure and customer support during those months.

Final thoughts

This project was a major milestone in my journey as a beginner data analyst. It taught me that data doesn’t need to be complex to be valuable, what matters is how well you clean, analyses and communicate insights.

If you are learning data analysis, I highly recommend starting with real-world dataset and building small projects like this. Every project moves you one step closer to mastery.

Thanks for reading.