📘 Product Feature Usage Intelligence Dashboard

End-to-End Business Analyst Project (Python • Streamlit • RFM • KMeans • Plotly)

---

📌 1. Project Overview

This project simulates a real SaaS product analytics workflow, where we analyze how users interact with different product features over time.

The system provides:

• Feature usage metrics
• User engagement analytics
• RFM-based behavioral scoring
• KMeans usage segmentation
• A polished multi-page Streamlit dashboard

This project is designed for Business Analyst / Product Analyst / Data Analyst roles and is fully runnable locally.

---

📌 2. Problem Statement & Business Context

Modern digital products generate large volumes of event-level data (clicks, searches, feature interactions). However, companies often struggle with:

• Understanding which features are actually used
• Identifying “power users” vs “at-risk” users
• Measuring adoption and engagement
• Making data-driven product decisions

This project solves that by simulating:

• Feature-level adoption
• Usage trends
• RFM-based engagement scoring
• Usage-based user segmentation
• Monitoring active vs inactive users

Product Managers, Growth Teams, and Analysts use dashboards like this to:

• Prioritize the roadmap
• Improve user retention
• Identify adoption gaps
• Personalize engagement or marketing campaigns

---

📌 3. Tech Stack

🧠 Programming & Analytics

🖥️ App / UI

🔧 Utilities & Engineering

---

📌 4. Data Description

Synthetic dataset simulates clickstream-like product usage:

Raw Data Columns

Column	Description
user_id	Unique user ID
signup_date	When the user joined
event_date	Date of feature usage
feature_name	Feature used (Search, Dashboard, API, etc.)
events_count	Number of actions for that feature/day

Processed Data Columns

Column	Description
last_event_date	Last active day
active_days	Days user engaged with product
total_events	Total feature interactions
recency	Days since last activity
frequency	Number of active days
monetary	Usage intensity score
cluster	KMeans behavioral segment

---

📌 5. Architecture

Synthetic Feature Usage Data
          ↓
Preprocessing (clean, transform, aggregate)
          ↓
RFM Feature Builder (recency / frequency / monetary)
          ↓
KMeans Segmentation Model (Power, Regular, At-Risk, Dormant)
          ↓
Streamlit Multi-Page Dashboard
          ↓
Insights for Product & Business Decision-Making

---

📁 Folder Structure

Product-Feature-Usage-Intelligence/
│
├── app/
│   ├── Home.py
│   └── pages/
│       ├── 1_Overview.py
│       ├── 2_Feature_Usage.py
│       └── 3_RFM_Segments.py
│
├── data/
│   ├── raw/
│   └── processed/
│
├── models/
│   └── kmeans_rfm.pkl
│
├── scripts/
│   ├── generate_synthetic_data.py
│   ├── preprocess_data.py
│   ├── build_rfm.py
│   └── train_model.py
│
├── src/
│   ├── preprocessing.py
│   ├── rfm.py
│   └── viz.py
│
├── tests/
│   └── test_predict.py
│
├── screenshots/
│   ├── overview.png
│   ├── feature_usage.png
│   └── rfm_segments.png
│
├── .gitignore
├── requirements.txt
└── README.md

---

📌 6. How to Run (Step-by-step)

1️⃣ Clone the repository

git clone https://github.com/girishshenoy16/Product-Feature-Usage-Intelligence
cd Product-Feature-Usage-Intelligence

2️⃣ Create a virtual environment

python -m venv venv
venv\Scripts\activate

3️⃣ Install dependencies

python.exe -m pip install --upgrade pip
pip install -r requirements.txt

4️⃣ Generate synthetic data

python scripts/generate_synthetic_data.py

5️⃣ Preprocess / build features

python scripts/preprocess_data.py
python scripts/build_rfm.py

6️⃣ Train KMeans model

python scripts/train_model.py

7️⃣ Evaluate KMeans model

python src/evaluate_model.py

8️⃣ Testing the project

pytest -v

9️⃣ Start the dashboard

streamlit run app/Home.py

Your dashboard opens at: 👉 http://localhost:8501

---

📌 7. Dashboard Walkthrough (with screenshots)

🏠 Home Page

Simple project overview and navigation.

---

📌 Overview Page

Shows high-level KPIs:

• Total users
• Avg events per user
• Avg active days
• RFM summary per cluster

---

📈 Feature Usage Page

Includes:

• Total usage per feature
• Interactive filters (date range, feature selection)
• Feature-wise time-series trends

---

🧮 RFM Segments Page

• 3D RFM scatter plot (Recency–Frequency–Monetary)
• Cluster distribution bar chart
• Insightful cluster interpretation

---

📌 8. Key Insights & Results

Some example insights (your numbers may differ):

🔹 Feature Usage

• "Dashboard" is the most used feature.
• API & Integrations have lower adoption (good candidate for UX improvement).

🔹 Engagement (RFM)

• Power users show low recency and high frequency/monetary.
• At-risk users show high recency with declining frequency.

🔹 Segments

Model successfully separates users into:

• Cluster 0 – Power Users
• Cluster 1 – Regular Users
• Cluster 2 – At-Risk Users
• Cluster 3 – Dormant Users

These are directly usable for:

• Re-engagement campaigns
• Feature onboarding
• Product roadmap planning

---

📌 9. Future Work

To expand this into a more advanced product analytics suite:

• 📉 Churn prediction model
• 🧩 Cohort retention heatmaps
• ⚡ Real-time usage ingestion (Kafka → DB → Dashboard)
• 📊 Feature correlation matrix (which features drive stickiness?)
• 🧭 User journey funnel visualization
• 🚀 Deploy dashboard to Streamlit Cloud / Render / AWS