Swiggy-customer-Behaviour-and-Sentiment-analysis

🍽️ Swiggy Customer Behaviour and Sentiment Analysis

A full-stack analytics solution designed to explore customer behavior, delivery performance, and sentiment trends across Bangalore’s food delivery ecosystem. This project analyzes 200,000+ orders to uncover insights on cuisine preferences, delivery ratings, payment methods, and customer sentiment—empowering operations teams and marketing strategists with predictive insights and interactive dashboards.

🚗 GitHub Project Repository

🔗 Click to view Swiggy-customer-Behaviour-and-Sentiment-analysis

🧠 Project Overview

Customer satisfaction in food delivery hinges on speed, quality, and experience. This project delivers an end-to-end analytics platform that enables:

📊 Behavioral analysis across demographics and locations
😊 Sentiment tracking and prediction
🕒 Delivery performance benchmarking
📍 Cuisine and payment preference mapping

🎯 Key Objectives

Clean and preprocess customer order data
Engineer features for delivery satisfaction modeling
Build classification models to predict delivery ratings
Deploy interactive dashboards for stakeholder decision-making

📁 Project Structure

File Name	Description
`swiggy_data.csv`	Raw dataset with 200K+ customer orders
`cleaned_swiggy.csv`	Preprocessed dataset with engineered features
`rf_model.pkl`	Trained Random Forest model for delivery rating prediction
`swiggy.sql`	SQL queries for filtering and aggregating customer data
`sqlconnect.py`	Python script for SQL database connection
`app.py`	Streamlit app for dashboard deployment
`swiggy customer behaviour and sentiment analysis.ipynb`	Jupyter notebook with EDA, modeling, and insights
`swiggy customer behaviour and sentiment analysis.pbix`	Power BI dashboard visualizing customer trends and satisfaction metrics

🧹 Data Preprocessing

Converted Order_Date to datetime format
Extracted time-based features (hour, day, month)
Encoded categorical variables (Cuisine_Type, Payment_Method, Sentiment, Gender)
Removed duplicates and handled missing values
Normalized Order_Value and Delivery_Time_Min for modeling

📈 Exploratory Data Analysis

📍 Cuisine popularity across Bangalore areas
🕒 Delivery time distribution by age and gender
💳 Sentiment trends by payment method and food rating
😊 Correlation between delivery rating and customer satisfaction
📅 Time-of-day impact on delivery performance

🤖 Modeling Approach

Target Variable: Delivery_Rating
Algorithms Used: Random Forest, Logistic Regression
Evaluation Metrics: Accuracy, Precision, Recall, F1 Score
Top Features: Delivery_Time_Min, Order_Value, Cuisine_Type, Sentiment, Customer_Age