Embedded Files
Customer Segmentation Eagle National Bank
User : Director of Finance and Strategy
Background : The retail industry is becoming more competitive. An effective and customer-focused business strategy is needed. A customer-centric business strategy will foster a strong relationship between customers and the bank, target the right audience, facilitate relevant product development, and provide tailored services.
Method and Tools : Customer profiles are segmented using K-Means and RFM for both regular and luxury loans. The execution was done using Python and visualized with Tableau.
Result : Potential Loyalist (447), At Risk (175), and Can't Lose Them (171) are customer segments that have quite large numbers, showing the potential for a large influence on business success. Lost (60) and About To Sleep (49) are segments that need special attention to prevent customer loss. Build relationships through on-boarding support and provide early success to customers and send relevant communications and add value through resources and product recommendations.
Recommendation of the Best Brands for Luxura - Hypothesis Testing and Linear Regression using Spreadsheets
User : CEO of Luxura
Background : Luxura needs a recommendation for a brand that can yield the best results for their business.
Method and Tools : We will use hypothesis testing analysis to identify brands that have a significant impact and regression analysis for recommending the promotional price for Adibi.
Result : The recommended brand for their business is Cellina based on Hypothesis Testing and DA. In Linear Regression, a negative relationship was found between Adibi's promotions and Adibi's Value of Order.
Statistic SpreedSheet.pdfA/B Testing E-Commerce Website
User : Product Managers
Background : The company has developed a new web page in order to try and increase the number of users who "convert," meaning the number of users who decide to pay for the company's product. The goal is to work through this notebook to help the company understand if they should implement this new page, keep the old page, or perhaps run the experiment longer to make their decision.
Method and Tools : A/B Testing chi-square using Python.
Result : The P-value of 0.2291is greater than the general significance level (usually 0.05), indicating that there is not enough statistical evidence to reject the null hypothesis.
There is no strong evidence to supportthe introduction of new pages. Therefore,the current recommendation may be to retain the old page or consider re-evaluating the new page design or elements before making any significant change decisions.
Customer Retention using Cohort Analysis
User : Marketing Manager.
Background : In the advancing digital era, online retail companies have become major players in the trading industry. Alongside the rapid growth of e-commerce, competition in the online market has become increasingly fierce. Therefore, online retail companies need to have a deeper understanding of their customers' behavior to remain relevant and sustainable in this industry. Customer retention rate has become one of the key indicators of success for online retail companies. The retention rate measures the extent to which a company can retain existing customers over a specific period. By understanding the factors influencing customer retention, companies can take strategic steps to improve retention rates and build strong relationships with customers.
Method and Tools : The method is Cohort Analysis using Python.
Result : The highest number of users made their first transactions in December 2010 (942). This user cohort also exhibited a significant rate of repeat transactions in November (50%) compared to other months. That cohort of users demonstrated the highest loyalty by continuing to transact in the subsequent months after their initial transaction, with a retention rate exceeding 30%. Unfortunately, a portion of customers did not make repeat transactions, evident from a retention rate not exceeding 50%. Of concern is the fact that the retention rate in April 2014 is the lowest for all user cohorts compared to other months.
Bank Loan Approval Prediction Using Logistic Regression
User : Risk Manager.
Background : The need to improve efficiency and accuracy in the loan approval process is essential. Facing the growing demand for loan services, the bank encounters challenges in managing an increasingly large volume of customer data while also enhancing the loan approval rate without compromising the quality of credit decisions.
Method and Tools : The method is Machine Learning Logistic Regression using Python.
Result : The model shows good accuracy on both training and test data, with minimal differences between the two. This suggests that the model may have the ability to generalize well to new data. Note that the model performs better for the rejected class (94%) than for the accepted class (53%). This indicates an imbalance in the data, where the model is better at identifying unqualified customers than qualified ones.
Credit Card Approval.pdfData Vizualization Tableau - Steam Games
User : CEO
Background : The CEO needs a dashboard to monitor game trends and company business performance.
Method and Tools : Tableau.
Result : Dashboard Monitoring. The dashboard includes a Line Chart comparing Total Price and Total Player Hours, Game Categories, Top 3 Games, Average Playtime, Average Purchase, and Total Playtime.
Business Problem
Background : As a data analyst at a large company called TFS Bank Ltd, I provide insight into credit risk and how to implement financial strategies to maintain healthy financial performance. One of my duties to find clues or patterns in suggesting potential credit risks.
Method : DARCI
Result : Metrics Recomendation
Understanding Business Problem.pdfPage updated
Google Sites
Report abuse