Analyzing data from e-commerce company databases using Python.
This was a database from the sales and inventory of an ecommerce store. There were a total of 8 tables in the database having different relationships with one another in the schema. There was data at the order level, customer demographics, seller demographics, product details, locations, item descriptions. No. of records was over 110k.
- pandas
- numpy
- matplotlib
- datetime
- scipy
- sklearn
- seaborn
- customer acquisition and retention for the company
- monthly revenue trends
- seasonality in sales
- testing hypotheses such as 'level of detail in product description affects sales quantity of a product'
- top n popular, least popular, costliest and cheapest product groups
- assigning product of the month based on popularity
- finding state wise preferences of shopping
- tracking delivery times and testing any correlation with review ratings
- customer segmentation (based on the amount of spending they prefer)
- joining tables in various ways for different purposes
- preprocessing, cleaning, typecasting, missing value treatment
- feature creation
- aggregations
- visualizations
- modelling and hypothesis testing
Key results and inferences in notebook alongside.