Palak Gupta👋
Turning data into insights with my Strategic Data Analysis
Portfolio Project 4:
Swiggy Data Analysis
Overview
The Swiggy Data Analysis project aimed to explore food delivery patterns, customer preferences, restaurant performance, and operational efficiency using real-world Swiggy data. The objective was to derive actionable business insights for restaurant partners and delivery management to enhance customer satisfaction and optimize logistics.
Research: Initial research focused on understanding how food delivery platforms operate, with special attention to order frequency, delivery times, popular cuisines, pricing, and customer ratings. The analysis targeted identifying demand trends, peak hours, and the impact of offers/discounts on order volume.
Information Architecture:The dataset was structured around key variables: order ID, customer ID, restaurant details, order time, delivery time, cuisine type, ratings, and discount codes. Data was cleaned, transformed, and categorized for better readability and analysis.
Wireframing and Prototyping:A Power BI/Streamlit dashboard prototype was created to visualize metrics such as average delivery time, top-performing restaurants, customer churn rates, and order heatmaps by time and location.
Challenges
Incomplete and Noisy Data:
- Challenge:Missing delivery timestamps, inconsistent restaurant names, and duplicate orders.
- Solution: Applied data wrangling techniques to impute missing values, deduplicate entries, and standardize categorical fields like restaurant and cuisine names.
Temporal Analysis:
- Challenge:Capturing peak order times, frequency patterns by day/hour, and delivery delays.
- Solution: Converted timestamps to datetime objects and extracted features like hour, weekday, and delivery duration. Visualized patterns using time-series plots.
Customer Segmentation:
- Challenge:Understanding diverse customer behaviors for targeted marketing.
- Solution: Used RFM (Recency, Frequency, Monetary) analysis and clustering (K-Means) to identify loyal, casual, and at-risk customers.
Geolocation Mapping:
- Challenge:Analyzing performance by zones or neighborhoods.
- Solution:Mapped delivery locations using geospatial plotting libraries to identify high-demand areas and delivery hotspots.
Results/Conclusion:
The analysis uncovered that evening hours and weekends drove the highest order volumes, with biryani and fast food as top cuisines. Delivery delays correlated with rain and traffic zones. The customer segmentation model provided insights for loyalty campaigns and personalized offers. Restaurant partners benefited from performance metrics, including customer rating trends and average order value. The project added real business value by providing data-backed recommendations for operational improvements and marketing strategies.
