Customer Quality segmentation

Data source is excluded from this repository.

Project Objectives:

The primary goal of this project is to measure the quality of newly acquired customers (i.e who have at least 1 successful order) for an e-commerce company using data analysis and machine learning techniques. The key objectives to achieve this goal are as follows:

1. Define the target variable that will be used to measure customer quality.
2. Predict the quality of new customers.
3. Segment customers to identify distinct customer groups by using your output.
4. Recommend strategies to improve customer loyalty.

Data:

market_order:

• client_id: The unique identifier assigned to a user when they register on the app. This ID is used to track the user's activities and purchases on the platform.
• order_id: The unique identifier of the particular order.
• checkoutdate: The date when the user checked out or finalized their purchase.
• createdate: The date when the basket was created or when the first product was added to it.
• city_id: The unique identifier for the city of the user.
• devicename: The name or model of the device used by the client to make the purchase.
• addressname: Name or label associated with the delivery address (e.g., "Home," "Work").
• distance: The distance between the warehouse and the delivery address.
• aandm: Financial metrics associated with advertising and marketing costs.
• basketvalue: The total value of the products in the basket.
• chargedamount: The total amount charged to the user including all fees and discounts.
• deliveryfee: Fee charged for delivery.
• discountamount: Amount of discount given on the basket.
• suppliersupport: Financial support provided by the supplier, if any.
• thirdpartysupport: Financial support provided by third parties, if applicable.
• ratedate: The date when the user provided a rating for their purchase.
• rateisskipped: Indicates if the user skipped giving a rating.
• rating: The rating given by the user for their purchase experience.
• ratereason: The reason associated with the rating, usually when the rating is negative or below a certain threshold.
• deliverdate: The date when the products were delivered to the user.
• warehouse_id: Unique identifier for the warehouse from which products were dispatched.
• hex_id: Hex ID related to the geolocation coordinates of the delivery location.

app_open:

• client_id: The unique identifier of the user who opened the app.
• appopendate: The timestamp when the user opened the app.

promo_detail:

• order_id: The unique identifier of the particular order.
• promo_id: Unique identifier for the promotion.
• responsibledepartment_id: Department responsible for the promotion.
• promoobjective: Objective or goal of the promotion, e.g., increase sales, attract new customers.

product_detail:

• order_id: The unique identifier of the particular order.
• _pos: Position or rank of the product, possibly in search results or listings.
• count: Quantity of the product.
• product_id: Unique identifier for the product.
• price: Price of the product.
• subcategory_id: Unique identifier for the subcategory of the product.
• brand_id: Unique identifier for the brand of the product.
• mastercategory_id: Unique identifier for the main or master category of the product.