UK Giftware Wholesale Retailer: Customer Insights & Recommendations

Table of Contents

1. Project Overview
2. Key Questions Explored
3. Tools and Technologies
4. Process Breakdown
5. Considerations & Limitations
6. Conclusions & Recommendations
7. Additional Information
8. Data Source

Project Overview

This project centers on a UK-based gift and giftware wholesale retailer seeking to leverage data analytics to enhance their business strategies. Historically, the company has struggled to consistently track and analyze its numbers. With a year's worth of transaction data now available, we aim to uncover valuable customer insights, make actionable recommendations, and estimate customer lifetime value (CLV) using machine learning.

Key Questions Explored:

1. Customer Insights: What patterns and behaviors can we identify from the transaction data?
2. Strategic Recommendations: How can the company leverage these insights to improve its operations?
3. CLV Estimation: Using a machine learning model, what is the estimated customer lifetime value (CLV) for the next 3-6 months?

Tools and Technologies

• Python (Jupyter Notebooks): For data cleaning, exploratory data analysis (EDA), and machine learning.
• MySQL Workbench: To run queries and validate calculations related to customer behavior.
• Tableau: For visualizing customer purchase frequency, revenue trends, and other key metrics.

Process Breakdown

Exploratory Data Analysis (EDA)

• Conducted in Jupyter Notebooks, where I cleaned and prepped the raw transaction data.
• Focused on understanding customer behaviors, such as purchase frequency and monetary value.

Database Exploration

• Exported the clean data to MySQL Workbench for further exploration.
• Queried the database to calculate metrics like frequency, recency, and monetary value, providing a solid foundation for CLV modeling.

Customer Lifetime Value Prediction

• Leveraged the lifetimes Python library to predict future purchases and estimate CLV for the next 6 months.
• These predictions offer a glimpse into future revenue potential, aiding in strategic planning.

Data Visualization

• Created visualizations in Tableau to explore revenue distribution, customer purchasing patterns, and more.
• These visualizations help to communicate the data's story effectively to stakeholders.

Considerations & Limitations

While this analysis provides meaningful insights, there are a few limitations to note:

• Lack of Profit and Cost Data: Insights are based solely on revenue data; profit margins and acquisition costs are not accounted for, limiting the depth of financial analysis.
• Limited Historical Data: With only one year of data, it's challenging to determine the 'true' customer lifetime value. Assumptions were made based on the available data.
• Incomplete Marketing Information: Without data on marketing activities, it's difficult to fully understand the drivers behind certain sales spikes.

Conclusions & Recommendations

• Seasonal Purchasing Behavior: 50% of revenue comes from customers making up to 5 purchases per year, aligning with seasonal buying patterns typical of wholesale buyers. However, an unexpected sales spike in November suggests further investigation into potential external influences like marketing pushes.
• Customer Retention: Over 60% of customers made repeat purchases, indicating strong customer relationship management practices. This is a key strength to build upon.
• Pricing Strategy: Consider reviewing the pricing structure and wholesale minimums. Many products were sold at wholesale prices with low quantities, leading to higher return rates. Implementing a minimum order quantity or adjusting retail pricing for smaller orders could mitigate this.
• Growth Opportunities: Exploring tradeshow participation during February and July could capitalize on seasonal buying trends, potentially boosting early-year sales.

Additional Information

This project was developed as part of my final project for the Data Analyst bootcamp at Ironhack. You can view the slideshow presentation here.

With a background in project management, I’ve always focused on helping small businesses improve their operations. Through this bootcamp, I've deepened my expertise, enabling me to craft data-driven proposals and strategies that drive impactful change.

Data Source

The original dataset was sourced from UC Irving Machine Learning Repository here.