I am captivated by the potential of data analytics, a field that empowers me to unravel complex datasets and derive insights that shape strategic decisions. This fuels my commitment to crafting impactful projects. I am actively seeking collaboration opportunities with like-minded individuals or organisations to explore innovative analytics projects. I am also enthusiastic about embracing new learning opportunities to further refine my skills and broaden my impact in this dynamic domain.
This challenge involved analysing a real banking transaction dataset (early 2023 to May 2025) with 21.64K transactions and 8,009 unique customers.
An in-depth analysis of the data provided some important findings:
-
Mobile and online channels are the most popular in terms of volume and fee revenue, which means that customers have fully accepted digital banking. However, the branch channels still record a greater efficiency of fee per transaction, implying that while the digital growth is high in terms of scale, it is yet to be fully streamlined in terms of voluntary fees.
-
The loan and credit card products are the main sources of fee revenue, and high-value everyday accounts like savings and checking do not yield many fees. This underscores excessive dependence on credit-related revenues and fines as opposed to basic transactional services.
-
The retention of customers is reduced drastically: cohort retention decreases to 710% in a few months. The bank is very good at customer acquisition and has challenges in long-term engagement, with an average of only 2.7 transactions per customer in a duration of more than two years.
-
High Customer Score segments (Good and Excellent) are bringing in significantly less fee revenue than expected, and customers with lower scores take up most of the penalties. As a result, there is a significant amount of upsell that is being missed with the most trustworthy customers.
An analysis of an ad campaign efficacy for an agency overseeing 20 client companies from 2022 to 2026, as part of my participation in The Analyst Challenge.
Leveraging DAX and SQL, I transformed and analysed the data, creating custom measures like profit, CPC, CPA, and more. A key highlight was employing cohort analysis in SQL to evaluate spend quartiles and campaign ID returns, uncovering critical trends.
In the data visualization process, I discovered that despite a remarkable 316% ROI, higher spending didn’t correlate with higher profit or conversions, as lower spending even yielded a higher CTR, pointing to spending inefficiencies. The data showed consistent performance across gender, customer segment, age, and location, with broad accessibility, though Pinterest lagged with the highest spend yet least profit and high CPA. As an agency managing ads for clients, I discovered 20 companies in 2022, a number that even declined by 2026, with most campaigns launched in 2022. This suggests that, despite strong metrics, productivity has waned, a trend worth addressing.
I designed a Power BI dashboard for a sample global electronics retailer to optimize business operations. Drawing from four datasets - Sales (transaction details), Products (product information), Customers (demographics), and Stores (location data).
This reports predominantly tracks critical KPIs, offering stakeholders a multi-faceted perspective to inform data-driven decisions. I utilised DAX in Power BI to construct RFM analysis, a vital and insightful approach for retail to segment customers based on Recency, Frequency, and Monetary value, alongside cohort analysis to evaluate long-term performance trends. Additionally, I employed parameters to enhance the dashboard's dynamic and interactive experience. This project served as my graduation deliverable for a comprehensive 4-month data analytics course.
An interactive dashboard on Looker Studio to analyse sales, customer engagement, and product performance for an e-commerce entity. My process began with meticulous data cleaning in Excel, followed by importing and visualising data from ecomsales, customer, product, and region sheets, delivering a comprehensive view of business dynamics.
An in-depth analysis of the universal Adventure Works sales dataset, spanning 2020 to 2022. My work uncovered significant trends in revenue generation, product efficacy, regional sales patterns, and customer demographics, providing a solid foundation for strategic decision-making.
- Power BI: For data modelling, DAX scripting, and dashboard creation
- Looker Studio: For visualising ecommerce and sales metrics
- Excel: For initial data cleansing and transformation
- SQL: For advanced querying and cohort analysis
- Figma: For designing intuitive layouts and visualisations