An Exploratory Data Analysis on e-Commerce events

This dataset contains 900K detailed users events from an electronics store e-Commerce website

Image credits - rupixen.com from Unsplash

About the data

In this analysis we will use the "e-Commerce events history in electronics store" data by Michael Kechinov on Kaggle from the REES46 Marketing Platform.

The Kaggle dataset is in .csv file format having 900K row and 9 columns.

Each row in the file represents an event. All events are related to products and users. Each event is like many-to-many relation between products and users.

Below are description of 9 columns

1. event_time: Time when the event occured (in UTC) in the format "2019-10-01 00:00:04+00:00"
2. event_type: A single event either "view", "cart" or "purchase"
3. product_id: A unique code for each product
4. category_id: A unique code for each product category
5. category_code: A '.' separated text with names of categories and subcategories "computers.network.router"
6. brand: Brand name of the product "sony"
7. price: Price of the product in decimals "31.90"
8. user_id: A permanent unique ID per shopper "1515915625519388267"
9. user_session: A temporary ID per session "LJuJVLEjPT"

Events in the column event_type can be:

1. view - a user viewed a product
2. cart - a user added a product to shopping cart
3. remove_from_cart - a user removed a product from shopping cart
4. purchase - a user purchased a product

What does each column tell us

At time event_time and during session user_session, user user_id purchased (event_type is "purchase") a product product_id of brand brand in category category_code of price price.