AI-POWERED CUSTOMER SEGMENTATION (Advanced)

Project Overview

This advanced version of customer segmentation goes beyond standard clustering. It enriches raw marketing data through advanced feature engineering, PCA-based visualization, and AI-based prediction of high spenders — offering both exploratory and actionable intelligence.

It simulates a marketing analyst’s toolkit for uncovering latent patterns in customer behavior and empowering decision-makers with clear visual insights.

Objective & Business Context

In modern customer-centric industries, raw demographics and spending data often lie unused in flat tables. The project addresses these gaps by:

• Segmenting customers based on nuanced behavioral signals like web purchases, recency, tenure, and lifetime spend.

• Visualizing these segments in an interpretable, decision-friendly form.

• Predicting high-spending customers using logistic regression to aid targeted marketing.

Such segmentation is crucial in retail, e-commerce, telecom, and banking, where user diversity and campaign efficiency directly affect revenue.

Business Value and Real-World Scope

• Marketing ROI Boost: Use intelligent segmentation to tailor campaigns and reduce customer churn.

• Predictive Customer Value: Pinpoint who’s likely to spend more, and when.

• Behavior-Based Targeting: Move beyond basic demographics and toward lifecycle-driven campaigns.

• Practical Deployment Ready: Streamlit dashboard offers stakeholder-ready insights in a single screen — no code required.

• Scalable Approach: Easily extended to include RFM, CLTV, churn prediction, or multi-touch attribution models.

Implementation Flow

• Dataset: marketing_campaign.csv

• Source: Public retail campaign dataset on Kaggle

• Key Fields:

  • Income, Age, TenureDays, TotalSpend

  • Product purchase data: MntWines, MntFruits, MntGoldProds, etc.

  • Behavioral engagement: NumWebPurchases, Recency, NumWebVisitsMonth

  • Demographics: Education, Marital Status, Kidhome, Teenhome

Dataset Overview

Data Preprocessing

• Cleaned nulls, encoded types.

• Derived new columns:

  • Age (from year of birth)

  • TotalSpend (sum of product category spend)

  • TenureDays (from customer registration date)

  • TotalChildren (combined from Kidhome + Teenhome)

Exploratory Data Analysis (EDA)

• Visualized distributions (e.g. income, recency)

• Plotted correlation matrix for behavioral patterns

• Analyzed categorical fields like education/marital status

Clustering (Unsupervised Learning)

• Used StandardScaler + KMeans on selected features

• Segmented customers into 4 clusters (configurable)

• Elbow method plotted to select optimal k

• Pair plots and boxplots showed intra-cluster differences

Dimensionality Reduction (PCA)

• Applied PCA for visualization (PC1, PC2)

• Cluster overlays on PCA scatterplot

• Explained variance tracked to ensure interpretability

High-Spender Classification (Supervised Learning)

• High-spender logic: MntWines + MntMeatProducts > 500

• Logistic regression trained on:

  • Income, Recency, Web Purchases, TotalSpend, Age

• Generated predictions (SpendingPrediction) + evaluation metrics

• Confusion matrix and classification report plotted

Streamlit Dashboard

• Interactive filters, data previews

• EDA plots, PCA plots, clustering visuals

• Prediction summaries + downloadable results

Output File

• customer_segments.csv – Includes:

  • Final cluster labels

  • Spending prediction (0 = low spender, 1 = high spender)

  • PCA values (PC1, PC2)

  • All relevant engineered features

🟡 This file is not saved locally by default. It is offered for download from the dashboard using st.download_button().

Visual Dashboard Overview

Histograms (Income, Recency, Total Spend, Age, Tenure)

Show the distribution of key traits like income and engagement. Help identify outliers and customer spread.

Boxplots (Income, Spend by Cluster)

Reveal spending and income variation across clusters. Useful for comparing financial value of each segment.

PCA Scatter Plot

Displays customers in a 2D space by similarity. Clearly shows how well clusters are separated.

Correlation Heatmap

Highlights relationships between behaviors and demographics. Helps spot patterns like high income = high spend.

Cluster Summary Cards

Quick insights into each group’s income, age, and spending. Ideal for creating buyer personas.

CSV Export

Lets users download the enriched, segmented data. Enables deeper analysis in Excel or BI tools.

✅ customer_segments.csv – Exported clustered + predicted customer data
✅ Streamlit dashboard – Interactive controls and visual analytics
✅ Logistic Regression – High spender classification with evaluation metrics
✅ Documentation & Codebase – Fully structured for reuse or extension

Key Deliverables

Tools and Libraries Used

• pandas – Data manipulation, cleaning, feature engineering, and CSV handling

• scikit-learn – PCA, StandardScaler, KMeans clustering, Logistic Regression classifier

• matplotlib, seaborn – Visual exploration: histograms, heatmaps, boxplots, PCA plots

• Streamlit – Interactive dashboard interface with filters, plots, and CSV export

• VS Code Terminal – Running scripts, managing environments, debugging

• RFM Segmentation: Incorporate Recency-Frequency-Monetary scoring for lifecycle marketing.

• CLTV Prediction: Predict long-term customer value using XGBoost or deep learning.

• Recommendation Systems: Suggest product bundles based on cluster behavior.

• Automation: Add auto-retraining scripts using Airflow or CRON.

• Personalized Reports: Export cluster-wise PDF summaries using pdfkit or reportlab.

Possible Next Steps & Conclusion

Conclusion

This project exemplifies the transition from raw data → segmented insights → actionable predictions. It mimics the exact steps followed in enterprise CRM and campaign analytics workflows.

By combining unsupervised and supervised learning with powerful visuals, it delivers:

• Deep behavioral understanding

• Real-time interactivity for marketers

• Direct pathways to boost revenue through smarter targeting

As AI becomes more embedded in business intelligence tools, this solution stands as a modular, explainable, and production-ready asset for real-world deployment in any customer-facing domain.

Dive into the foundational concepts, algorithms, and real-world relevance behind this project. From machine learning principles to business strategy insights, this conceptual study bridges the gap between technical implementation and applied decision-making—helping you understand not just how it works, but why it matters.

Key Concepts

GitHub Repository

Want to dive deeper into how this project actually works?

We’ve made the complete codebase and resources available for you on GitHub

👉 Access the full repository here:

Whether you're a learner, recruiter, or collaborator — there's something for everyone.