Customer Segmentation and Marketing Strategy Optimization: A Data-Driven Approach

Info	Description
Project Objective	Customer Segmentation
Assignee	Ahmed Yousri Sobhi
Assignee's email	[email protected]
Department	Data Science

Table of Content

• Customer Segmentation and Marketing Strategy Optimization: A Data-Driven Approach
  • Table of Content
  • Objective
  • Project Structure
  • Setup Environment
  • Dataset Schema
  • Dataset Info
  • Project Flow
  • Analysis Report

Objective

Our goal is to use data insights to better understand our customers. By analyzing their demographics, behavior, and how they respond to our marketing efforts, we aim to create personalized strategies that resonate with different customer groups. Ultimately, we want to boost engagement, loyalty, and revenue by tailoring our approach to what our customers truly want.

Project Structure

.
├── cluster                # Clustering Model Notebook
├── data                   # Dataset files
|   ├── raw
|   ├── intermid
|   ├── output
├── nb_workspace           # Jupiter notebooks
├── docs                   # Documentation files which includes analysis report
|   ├── model              # Exported trained models
├── nb_workspace           # Jupiter notebooks
├── report                 # Reports files
|   ├── analysis_report    # Analysis report
|   ├── plots              # Figures files
|   ├── reports            # Reports HTML files
├── src                    # Source scripts files
|   ├── env                # Script file to create project directory environment
|   ├── tools              # Script files used for model pipeline

└── README.md

Setup Environment

Navigate to the project directory:

cd project_directory

Create a virtual environment (optional but recommended):

python -m venv venv

Install the project dependencies:

pip install -r requirements.txt

In case of initial setup, create required folders directory using environment script:

python ./src/env/setup_env.py

Dataset Schema

Feature	Description	Type
'ID'	Customer’s unique identifier	int64
'Year_Birth'	Customer's birth year	int64
'Education'	Education Qualification of customer	object
'Marital_Status'	Marital Status of customer	object
'Income'	Customer's yearly household income	int64
'Kidhome'	Number of children in customer's household	int64
'Teenhome'	Number of teenagers in customer's household	int64
'Dt_Customer'	Date of customer's enrollment with the company	timestamp
'Recency'	Number of days since customer's last purchase	int64
'MntWines'	Amount spent on wine	int64
'MntFruits'	Amount spent on fruits	int64
'MntMeatProducts'	Amount spent on meat products	int64
'MntFishProducts'	Amount spent on fish products	int64
'MntSweetProducts'	Amount spend on sweets products	int64
'MntGoldProds'	Amount spend on gold products	int64
'NumDealsPurchases'	number of purchase deals	int64
'NumWebPurchases'	Number of purchases made through the web	int64
'NumCatalogPurchases'	Number of purchases made through catalogs	int64
'NumStorePurchases'	Number of purchases made in physical stores	int64
'NumWebVisitsMonth'	Number of web visits in a month	int64
'AcceptedCmp1'	Whether the customer accepted Marketing Campaign 1	int64
'AcceptedCmp2'	Whether the customer accepted Marketing Campaign 2	int64
'AcceptedCmp3'	Whether the customer accepted Marketing Campaign 3	int64
'AcceptedCmp4'	Whether the customer accepted Marketing Campaign 4	int64
'AcceptedCmp5'	Whether the customer accepted Marketing Campaign 5	int64
'Complain'	Whether the customer has made a complaint	int64
'Z_CostContact'	Cost associated with contacting the customer	int64
'Z_Revenue'	Revenue associated with contacting the customer	int64
'Response	Customer's response to the company's offer	int64

Dataset Info

This dataset are divided into 4 categories:

• Customer's Information: [ID, Year_Birth, Education, Marital_status, Income, Kidhome, Teenhome, DT_Customer, Recency, Complain]
• Products: [MntWines, MntFruits, MntMeatProducts, MntFishProducts, mnSweetProducts, MntGoldProds]
• Promotion: [NumDealsPurchases, AcceptedCmp1, AcceptedCmp2, AcceptedCmp3, AcceptedCmp4, AcceptedCmp5, Response]
• Place: [NumWebPurchases, NumCatalogPurchases, NumStorePurchases, NumWebVisitsMonth]

Project Flow

The projects is splitted into two parts, where:

Dataset Exploring:

• The first part describes all the preprocessing, exploarity data analysis (EDA), and features engineering the dataset has gone through.
• Locatted at: nb_workspace/

Customers Segmentation:

• The second part describes the pipeline which the raw dataset is passed through, to the output of customers clusters, passing through all the preprocessing process [Data Cleaning and feature engineering].
• Located at: cluster/customer_segmentation.ipynb

All Steps and Code are documented and commented in the notebooks, so any contributor would simply go through the steps made.

Analysis Report

Included inside directory -> report/analysis_report/analysis_report.pdf