Data Chronicles Blog: Insights for Business Growth
I’m an enthusiast looking to analyze data and build predictive models with Python. Though I am still learning, I have experience with SQL and Excel.
On this blog, I share data projects that provide business insights. You’ll find examples where I collect, clean, and analyze datasets, spotting trends and patterns to inform decisions on marketing, customer service, operations, and more.
I also showcase any machine learning prototypes I build using Python and other tools. I aim to display analytical thinking in the business context.
Customer Churn (EDA)
In my role as an entry-level data scientist at one of the leading credit card companies in the U.K., I was tasked with supporting the company’s goal for the new business year: to retain and grow the customer base.
To achieve this, I began by analyzing customer churn to understand why certain customers left. The insights gained allowed me to identify gaps in the system and develop strategies to address these issues.
Next, I focused on reactivating churned customers, leveraging the valuable customer data we still had, as it’s often easier and more effective to re-engage those who have already interacted with our brand.
Additionally, I worked on developing strategies to better retain existing customers by enhancing their engagement with our services.
Lastly, I analyzed customer demographics, particularly age, to uncover insights that could inform targeted marketing strategies and ad campaigns for the marketing team.
This comprehensive approach not only identified the customer leakage but also aimed to drive future growth through more personalized retention and acquisition strategies. The analyses can be found here
Breast Cancer Prediction (using KNeighbors Classifier)
A general hospital faced a surge in diagnostic errors, missing many cancer cases. This led to delayed treatments and worsened patient outcomes. To tackle this, I developed a predictive model using K-Nearest Neighbors (KNN) with 3 neighbors. The focus was on improving recall to reduce false negatives and ensure early cancer detection. You may check out the analysis here
Car Sales Predictiction
Automotive Insights, a car dealership, analyzed data from 1,000 customer records to understand buying behavior and improve marketing strategies. The data included age, salary, gender, and purchase history. As the analyst, I developed models to predict car purchases. Key findings: more women bought cars than men. Older customers bought cars despite lower incomes, showing changing priorities with age. Salary was crucial; younger customers prioritized education and personal development. Explore more in my portfolio
House Prediction
The real estate agency, N_Move-in Realtors, gathered data from the previous year to analyze a newly developed area in Wellingborough, Northampton. To predict house prices, this data was collected to provide valuable insights to landlords seeking to determine the market value of their properties for potential buyers. As the newly appointed data scientist, my task was to develop a model capable of accurately predicting house prices based on key features such as house age, distance to convenience stores, and distance to the nearest MRT station. The detailed analyses
Salary Prediction
As a data scientist at a growing startup, one of the challenges is establishing a fair and consistent pay scale across all roles while considering employees’ years of experience. Currently, each staff member negotiates their salary independently, and the company lacks both a structured pay scale and defined job levels.
Currently, there’s no structured pay system in place, and all employees negotiate their salaries individually, leading to a lot of discrepancies and biases in offers to new hires.
The aim is to introduce a transparent pay scale that categorizes roles based on years of experience, ensuring that new hires are not unfairly compensated compared to existing employees. Kindly click here here for the detailed analyses.
Ice-Cream Sales Prediction
As the data scientist for an ice cream company, leveraging temperature data offers a strategic opportunity to enhance sales performance. By analyzing the relationship between temperature and ice cream sales, actionable insights can be derived to optimize marketing strategies, inventory management, and product offerings. Developing a predictive model based on temperature fluctuations enables proactive decision-making, resulting in increased sales revenue and improved customer satisfaction. The detailed analysis
Unleashing Academic Potential: A Data-Driven Approach
I developed a project that harnesses multiple linear regression modeling to predict student academic performance. By analyzing factors like study hours, previous scores, sleep patterns, and practice materials, I created a predictive model that can help educators better support their students’ success. This model allows inputting a new student’s data on these factors to estimate their future performance index. Explore the code and data to see how this approach provides insights into understanding and nurturing academic excellence.
Predicting Sales from Ad Campaigns: A Consulting Start-up’s Analysis
A consulting start-up plans to analyze its marketing efforts from the past quarter to better understand the relationship between marketing costs and sales revenue generated.
This analysis will enable the start-up to allocate its marketing budget more effectively, identify the most and least efficient marketing channels, and develop predictive models for estimating sales based on marketing spending in each channel. The key goals are to optimize return on marketing investment, focus future spending on high-performing channels, and quantify expected sales lifts from different marketing initiatives. Marketing Ad Analyses
Sales Prediction for Nova, a Pharmaceutical Company.
Nova spends $10 million per year on pharmaceutical advertising but lacks data on the sales impact of this spending.
Analysis revealed a positive linear correlation between Nova’s advertising costs and sales over time. By leveraging analytics and modeling, Nova can optimize future budget allocations to maximize returns and sales lift from its $10 million ad spend. However, advertising is only one sales driver - Nova must examine broader initiatives to drive growth. Nova’s Prediction Analyses
Predicting SAT scores of students using their GPAs (Ordinary Least Squared Analyses/Simple Linear Regression Analyses. A case study).
I was approached by an educational institute that needed to increase the success rate of its SAT training. From experience, they knew there should be a correlation between the GPA of students and their SAT. I was approached to do a simple equation that can be used to predict the scores of applicants based on their GPA scores. That way, they know who to admit and who not to, to increase the overall success rate of their training. Check out the detailed analyses here
eBay Cars: Python Analysis
Let’s imagine I’m a new data scientist who has been assigned to a project for eBay. As an online shopping company, eBay wanted to better understand the car products sold on their website. They likely wanted to know what types of cars were advertised the most by sellers using eBay. This could help them decide if they should encourage sellers to advertise more of certain car models.
eBay also probably wanted to see what types of cars were actually selling the most to buyers. This information could reveal if buyer demand matched up well with what sellers were offering.
As the data scientist on this project, I needed to collect and analyze the data on car advertisements and sales happening on eBay. I would dig into the data to see the most common car makes and models shown in listings and purchased. My analyses could provide business insights to improve eBay’s car marketplace for both buyers and sellers.
Specifically, here is a link to my full analyses for this eBay car data project that I was assigned as a new data scientist: analyses.
Date posted: 1/1/2024
General Python Codes
I recently started learning Python coding, specifically for machine learning. Python’s versatility and beginner-friendly syntax made it appealing.
As a machine learning newbie, I wanted to document my journey in picking up this valuable skill.
In my initial attempts, I’m practicing core concepts like variables, data types, functions and more. While basic, understanding these building blocks will allow me to progress to machine learning and AI down the road.
I find coding rewarding already, like piecing together something useful. With more Python fluency, I look forward to translating machine learning models and ideas into code.
It will be exciting to see what I can build as I continue ramping up my Python skills for machine learning and AI applications.
Below are some snippets as I get started with learning Python basic codes and visualization
Date posted: 10/12/2023
Facility Management SQL Analysis
As a data analyst for a facility management retail company, I leveraged SQL to analyze the store’s sales transaction data. My goal was to help the company gain insights to improve inventory planning, pricing strategies, and targeted marketing campaigns.
I wrote SQL queries to study essential sales metrics across product categories, customer segments, seasonal trends, and store locations. My analyses provided a better picture of customer purchasing behavior and demand forecasts. Here are the analyses
##### Date posted: 10/11/2023
DVD Rental SQL Analysis
As an analyst for a DVD rental store, I used SQL to study their transaction data. My goal was to find ways to improve inventory, pricing, and marketing.
I queried the data to reveal insights into:
- Top renting movie genres
- Customer patterns by membership type
- Seasonal rental trends
My SQL analysis gave me a better understanding of customer behaviors. This can guide inventory decisions, pricing changes, and targeted ads for retention and growth. Here are the analyses.
Date posted: 5/11/2023
Starbucks_SQL_ Project
I analyzed survey data for Starbucks London to help them understand customers. My goal was to improve customer loyalty and attract new customers to make more money.
I studied demographics to see who their customers are. My analysis gave useful information. Now Starbucks can make better choices to earn higher revenue.
I found ways Starbucks London can keep existing customers happy. I also suggested how to appeal to new customers. My data analysis offered the insights they need to boost sales. Here are the results.
Date posted: 20/11/2023
DMart Supermarket PowerBI Analysis
I recently completed a data analysis project using Power BI to evaluate the product performance of a supermarket chain. Here are the details.
Date posted: 30/10/2023
Certifications
Contact Details
Feedback is welcome! Connect via kojusoluwaolutade@gmail.com or my resume.