FLO Customer Segmentation & RFM Analysis 👟📊

📌 Business Problem

FLO, an OmniChannel shoe retailer, wants to segment its customers and define marketing strategies according to these segments. To this end, the shopping behaviors of customers will be defined, and groups will be formed based on these behavioral clusters using RFM Analysis.

🎯 Project Goal

• To understand customer behavior using Recency, Frequency, and Monetary metrics.
• To create actionable customer segments (e.g., Champions, Hibernating, At Risk).
• To generate specific customer lists for targeted marketing campaigns.

📂 Dataset Story

The dataset consists of information obtained from the past shopping behaviors of customers who made their last purchases in 2020–2021 via OmniChannel (both online and offline).

Variable	Description
master_id	Unique customer number
order_channel	Channel used for shopping (Android, iOS, Desktop, Mobile, Offline)
last_order_channel	Channel where the last purchase was made
first_order_date	Date of the customer's first purchase
last_order_date	Date of the customer's last purchase
order_num_total_ever_online	Total number of purchases on the online platform
order_num_total_ever_offline	Total number of purchases on the offline platform
customer_value_total_ever_offline	Total fee paid for offline purchases
customer_value_total_ever_online	Total fee paid for online purchases
interested_in_categories_12	List of categories shopped in the last 12 months

⚙️ Methodology: RFM Analysis

1. Data Preparation:

   • Handling missing values and outliers.
   • Creating new metrics: Total Order Number & Total Spending.
   • Date format conversions.

2. Calculating RFM Metrics:

   • Recency: How recently did the customer purchase? (Analysis Date - Last Order Date)
   • Frequency: How often do they purchase? (Total Transaction Count)
   • Monetary: How much do they spend? (Total Price)

3. Calculating RFM Scores:

   • Metrics are converted to scores from 1 to 5.
   • RFM_SCORE is derived by combining Recency and Frequency scores.

4. Segmentation:

   • Customers are mapped to segments like "Champions", "Loyal Customers", "Hibernating" based on regex mapping of their scores.

📊 Key Findings & Visuals

Here is the summary of the customer segments identified:

Insights:

• Champions: There are 1,932 customers who shop frequently and recently.
• Cant_Loose: 1,200 customers have high purchase frequency but haven't visited recently. They need immediate attention.
• Hibernating: A large portion (3,604) of the customer base is inactive.

🚀 Actionable Scenarios (Case Studies)

Case 1: New Women's Shoe Brand Launch

Objective: Target loyal customers (Champions, Loyal Customers) who are interested in the "WOMEN" category for a premium brand launch.

• Action: - Segments: Champions, Loyal_Customers
  • Category Interest: KADIN
  • Output: new_brand_target_customer_id.csv

Case 2: Men's & Children's Discount Campaign

Objective: Reactivate lost customers and reward new ones interested in "MEN" or "CHILDREN" categories with a 40% discount.

• Action: - Segments: Cant_Loose, Hibernating, New_Customers
  • Category Interest: ERKEK, COCUK
  • Output: discount_target_customer_id.csv

💻 How to Run the Project

1. Clone the repository:

   git clone [https://github.com/merttdmrr/FLO-RFM-Customer-Segmentation.git](https://github.com/merttdmrr/FLO-RFM-Customer-Segmentation.git)

2. Install the requirements:

pip install -r requirements.txt

3. Run the Jupyter Notebook.

🔗 Author

Mert Can Demir