I started with 115 Jumia product records. Before analysing anything, I needed to understand the quality of the data. During the cleaning stage, I found three records with invalid review values, so I removed them. The final dataset contained 112 products.
I used Python and Pandas to inspect and clean the raw Jumia dataset before moving into the analysis and dashboard stages.
The original dataset contained 115 product records. During the cleaning process, I investigated the review field and found that the source data represented review counts as negative values. I also identified three invalid records and removed them.
After cleaning, the dataset contained 112 product records.
I also created additional analytical fields, including:
Review count
Current price midpoint
Old price midpoint
Discount amount
Calculated discount percentage
Rating category
Discount category
Price category
After cleaning the dataset with Python, I loaded the cleaned data into PostgreSQL for structured analysis.
I used SQL to investigate the main business questions around:
Product discounts
Customer review activity
Product ratings
Product prices
Highly discounted products
Highly reviewed products
I also used SQL queries to calculate summary statistics and identify products that stood out based on their ratings, review activity and pricing.
This stage allowed me to move from simply cleaning the data to analysing it from a business perspective.
Excel Analysis and Pivot Tables
After completing the data cleaning in Python and the structured analysis in PostgreSQL, I moved the relevant data into Microsoft Excel to build the reporting and visualization layer.
I used Excel to organize the analysis into summary tables and pivot tables. This made it easier to compare products based on discounts, ratings, review activity and pricing.
The pivot tables helped me identify patterns such as:
Products with the highest discount percentages
Products with the highest review activity
Rating categories across the dataset
Discount categories across products
Full project details can be found on:
https://github.com/joelm69/jumia-data-analysis
Top comments (0)