🌍 TravelTide Customer Segmentation Project

This project simulates a real-world data challenge from the online travel industry.

As a Data Analyst at TravelTide, I worked on segmenting customers to support the launch of a personalized rewards program. The goal was to deliver data-driven recommendations based on user behavior and help the marketing team tailor offers for better customer retention.

---

🎯 Project Goal

Segment TravelTide customers based on website behavior to enable targeted perks and personalized marketing strategies.
Support the marketing team in delivering more relevant offers and boosting customer loyalty.

---

🔍 Methodology

🗂️ Data Sources Used

• PostgreSQL database with the following tables:
  • users: user demographics and home location
  • sessions: user activity and booking sessions
  • flights: flight details per session
  • hotels: hotel bookings per session

🧠 Key Steps in the Segmentation Process

1. Exploratory Data Analysis (EDA):

   • Inspected all tables and data distributions
   • Joined tables to create a session-level base table
   • Cleaned and enriched the data for feature extraction

2. Feature Engineering:

   • Created behavioral metrics such as:
     • Total number of bookings
     • Hotel vs. flight booking ratio
     • Cancellation rate
     • Recency and frequency of activity
     • Discount usage
   • Aggregated features at the user level

3. Customer Segmentation:

   • Applied K-Means clustering
   • Labeled segments based on dominant behavioral traits
   • Assigned relevant perks to each segment

---

📊 Dashboard Design

The following dashboards provide interactive insights into user segments, preferences, and behavior.

1. 🎛 Segment Deep Dive Dashboard — Traveler Type

🔗 View Interactive Dashboard in Tableau Public

This dashboard compares customer segments based on:

• 💰 Average Lifetime Value (LTV)
• ✈️ Total Trip Volume
• 📅 Booking Frequency
• 🕒 Session Activity

Use the dropdown filter at the top to explore each traveler type individually. 📌 Filter by Traveler Type to analyze each segment individually.

2. 📈 KPI Dashboard – Value Metrics by Behavior

👉 View in Tableau Public

Includes:

• ADS per km
• Conversion Rate
• Discount Usage

---

👤 Behavioral Features Considered

• Average session duration
• Booking frequency (flights & hotels)
• Time between sessions
• Use of discounts and cancellations
• Total and average spending

---

🧠 Normalized Behavioral Profile by Traveler Type

This radar chart visualizes key behavioral differences between user segments:

• Session Duration
• Session Frequency
• Discount Usage
• ADS per km (Amount of money saved per kilometer)
• Conversion Rate

Each metric was normalized (min-max) to allow for fair visual comparison across different scales.

📊 Radar Chart Preview:

📁 Source Data:

• 📊 Radar_Chart_Template_-_Google_Sheets.xlsx
• 🔗 View in Google Sheets

📊 Key Findings

• Identified 5 distinct customer segments with unique travel behaviors
• Each group showed specific traits such as booking frequency, average spend, and presence of children
• Suggested perks are tailored to reflect user motivation and potential loyalty

---

🎁 Perk Strategies

Segment Name	Key Behavior Profile	Suggested Perks
Business Traveler	Frequent traveler, high km flown, short notice bookings	✈️ Priority boarding, 💼 Free lounge access
Dreamer	Low number of trips, inspired but inactive	🌍 Travel inspiration emails, 🎫 Welcome voucher
Family Traveler	Travels with children, medium booking rate	👨‍👩‍👧‍👦 Free child stay, 🛏️ Family package deals
Senior Traveler	Older age group, low frequency, steady behavior	🧓 Senior discounts, 📞 Personal booking assistant
Young Frequent Traveler	Frequent, independent travelers	🔄 Flexible cancellations, 🚀 Early access to deals

---

📈 Implementation Ideas

How segmentation can be implemented on the TravelTide platform:

• Personalized homepage content
• Behavior-based email campaigns
• Retargeting ads using segment tags
• Dynamic perks displayed based on user group

---

🔮 Next Steps

• A/B testing of perks across customer segments
• Integration with CRM for real-time personalization
• Track segment shifts over time
• Monitor KPIs: engagement, conversion, retention per group

---

🧰 Tools & Technologies

• SQL – Data extraction and joins
• Python – Feature engineering & clustering (K-Means)
• Tableau – Visualizations (bar charts, radar charts, scatter plots)
• Google Slides – Stakeholder presentation

---

📦 Deliverables

• ✅ README.md with full project description
• ✅ ERD diagram (view here)
• ✅ Feature matrix for clustering (users_with_segments.csv)
• ✅ Python notebook with clustering logic
• ✅ Tableau dashboard
• ✅ Executive presentation
• ✅ Final report (PDF)

---

📁 Directory Structure

TravelTide-Customer-Segmentation/
├── README.md
├── ERD.jpg
├── users_with_segments.csv
├── segmentation_notebook.ipynb
├── visuals/
│   └── dashboard_mockup.md
├── presentation/
│   └── Executive_Summary.pdf

---

📝 Summary

• Goal: Segment TravelTide users to deliver targeted perks based on behavior
• Data: PostgreSQL – users, sessions, flights, hotels
• Methods: SQL + Python (KMeans clustering) + Tableau
• Features: Bookings per session, discount usage, session time
• Outcome: 5 segments with tailored perks for personalization
• Next Steps: A/B testing, CRM integration, longitudinal tracking

---

📽️ Video Presentation

🎥 Watch the 5-minute presentation on Loom

🖥️ View the Google Slides

---