This project explores how nonprofit organisations can strategically use data to deepen donor engagement, increase retention, and maximise campaign impact. Using a custom-designed synthetic dataset that mimics real-world nonprofit operations, I applied advanced analytics techniques to uncover behavioural patterns, segment donors using a tailored RFM model, and evaluate the alignment between fundraising efforts and mission-driven outcomes.
Key components of the analysis included:
Recency-Frequency-Monetary (RFM) segmentation to classify donors into meaningful engagement groups such as Loyal, At-Risk, and High Value Potentials.
Retention trend analysis, revealing that donor loyalty averages ~30% annually, underscoring the need for smarter re-engagement strategies.
Campaign and impact alignment, showing how well (or poorly) funds raised translate into tangible outcomes.
Engagement response analysis, based on communication history and responsiveness across donor types.
The project concludes with seven stakeholder-ready recommendations that combine data-driven strategy with practical nonprofit experience - from personalising communication to building internal data capacity and automating donor journeys.
Tools Used: Python (Pandas, Seaborn, Matplotlib), Jupyter Notebook, Markdown, HTML/PDF Export
Role: Full project ownership - from dataset design and cleaning to analysis, reporting, and recommendations.
To access the case study report please click on the button provided below:
This project explores purchasing patterns in retail transaction data using Market Basket Analysis. By applying the Apriori algorithm, it identifies frequent item combinations and generates association rules based on support, confidence, and lift.
The analysis involves data cleaning, transforming transactions into basket format, and mining association rules. Visualisations such as scatter plots, bar charts, and network graphs highlight strong product relationships that can support strategies like product bundling, cross-selling, and personalised promotions.
Tools Used
Python (Visual Studio Code 2022, Jupyter Notebooks),
Pandas, Seaborn, Matplotlib, mlxtend, NetworkX
To access the case study report please click on the button provided below:
This project utilised the Superstore dataset from Kaggle to showcase a comprehensive data analytics workflow. Initially, the dataset was cleaned and modified—specifically by shortening product names - to ensure compatibility with Azure SQL Server. A structured relational database was then created in Azure SQL, with data imported and explored extensively using SQL Server Management Studio (SSMS).
Following database creation, exploratory data analysis (EDA) was carried out using Python, leveraging Visual Studio Code 2022 and Jupyter Notebooks to examine data trends, insights, and patterns. Throughout the project, effective documentation, version control, and collaboration practices were maintained using Git and GitHub. This work demonstrated practical skills in data cleaning, database management, SQL querying, Python-based analytics, and problem-solving, highlighting proficiency in essential tools and technologies for data-driven decision-making.
Tools Used
Azure SQL Server
SQL Server Management Studio (SSMS)
Python (Visual Studio Code 2022, Jupyter Notebooks)
Git and GitHub
To access the case study report please click on the button provided below:
This project explores sample retail sales data using SQL, Python, and Power BI to uncover key business insights. It analyses sales performance, customer behaviour, and purchasing trends while providing interactive visualisations for better decision-making.
Key analyses include:
✔ Monthly Sales Trends - Identifying revenue patterns over time
✔ Customer Segmentation - Grouping by age, gender, and lifetime value (CLV)
✔ Order Size Categorisation - Small, medium, and large order distribution
✔ Product Performance - Best-selling categories and revenue contribution
✔ Sales Forecasting - Predicting future trends using time-series models
The insights help businesses optimise marketing strategies, improve inventory management, and enhance customer engagement. The project is structured with SQL for data extraction, Python for analysis and machine learning, and Power BI for visualisation.
Tools Used
MySQL Workbench
Python (Jupyter Notebooks)
Power BI
Git and GitHub
To access the case study report please click on the button provided below:
Cyclistic is a bike-share program that provides users with an easy and efficient mode of transportation. The company aims to increase annual memberships by understanding rider behavior and identifying key differences between casual and member users. This study analyses data from Q1 2019 and Q1 2020 to uncover trends and provide actionable insights.
Tools and Technologies:
Excel
R (R Studio, R Shiny))
Python (Jupyter Notebooks)
Power BI
Git and GitHub
To access the case study report please click on the button provided below:
An interactive visual story of Fatal and Serious Injury Road Crashes in New Zealand including Chatham Islands Territory (2011-2020).
Population data used for the project was extracted from http://infoshare.stats.govt.nz/.
To access the case study report please click on the button provided below: