This project focuses on analyzing DoorDash delivery data to extract meaningful insights regarding dashers (delivery personnel), order deliveries, and revenue trends across various markets. I performed data exploration, cleaning, transformation, and in-depth analysis to uncover patterns in delivery times, order volumes, market performance, and customer behavior.
Food delivery services face challenges in optimizing delivery efficiency, managing dasher workloads, and improving customer satisfaction. High-demand periods lead to delays, inefficient resource allocation, and suboptimal revenue performance. This project aims to analyze delivery performance, order trends, and revenue contributions to identify key bottlenecks and provide data-driven recommendations for enhancing operational efficiency and maximizing profitability.
Here are resources related to the project:
- Project Documentation: Detailed project documentation outlining the approach, methodology, and results.
- SQL Queries: SQL scripts used for data extraction, cleaning, and analysis.
- Source Data: Raw data used for analysis, including store information and order details.
- Project Appendix: Additional project details and supplementary information.
- Dashboard File: Power BI dashboard used for data visualization and analysis.
- Data Exploration & Initial Analysis
- Data Cleaning
- Data Transformation
- Price & Duration Updates
- Categorization & Classification
- Column Removal & Adjustments
- Revenue & Performance Analysis
- Dasher Insights & Analysis
- Delivery Insights & Analysis
- Revenue Insights & Analysis
- Comparisons & Advanced Analysis
- Key Findings & Business Recommendations
- SQL Queries Used
- Tech Stack & Tools
- Retrieving all data from the
DoorDashtable for initial inspection. - Describing the table structure, checking column data types, and understanding the dataset.
- Identifying unique categories in
store_primary_category. - Handling missing values by identifying NULL categories.
- Grouping stores by category to analyze the distribution of stores.
- Removing unwanted data, such as eliminating stores categorized as
chocolate. - Checking for missing values in critical columns (
market_id,store_id,order_protocol, etc.).
- Creating new columns, including
city_name,subtotal_in_dollars,order_created_day_name, etc. - Modifying column data types (
actual_date,created_date,created_time,actual_time). - Updating columns with calculated values, such as converting
subtotalfrom cents to dollars.
- Converting prices from cents to dollars (
min_item_price,max_item_price). - Updating delivery durations, converting seconds to minutes (
estimated_order_place_duration_in_minutes). - Calculating total estimated delivery duration by summing various time components.
- Peak vs Off-Peak categorization (
Peak_Hourcolumn). - Item price segmentation into 'Low', 'Medium', and 'High' price categories.
- Time of day classification (
Morning,Afternoon,Evening,Night).
- Dropping unnecessary columns, such as
Item_Volume_Category. - Handling division by zero when calculating
revenue_per_item.
- Daily & hourly revenue trends to analyze peak revenue periods.
- Impact of delivery time on revenue based on time-of-day analysis.
- Performance metrics analysis, including dasher load factor, delivery duration, and outstanding orders.
- Market contribution to total revenue (highest and lowest-performing regions).
- Cuisines generating the most revenue and their popularity by order volume.
- Item pricing analysis by cuisine category.
- Dasher workload analysis across different markets.
- Impact of outstanding orders on delivery time to determine bottlenecks.
- Utilization patterns, identifying when dasher usage exceeds 80%.
- Idle dasher impact on overall delivery performance.
- Order density per dasher and peak-time order handling efficiency.
- Market-level analysis of average delivery times.
- Comparing actual vs estimated delivery times across all orders.
- Late delivery patterns and peak-hour delays by market.
- Peak hours for orders, identifying the busiest periods.
- High-value order delivery trends, analyzing if premium orders are delivered faster.
- Dasher load vs actual delivery times, identifying bottlenecks.
- Market performance comparisons based on KPIs.
- Predicted vs actual delivery time analysis, evaluating estimation accuracy.
- Factors influencing late deliveries, including order size and dasher availability.
- Top markets and cuisines ranked by revenue, order volume, and efficiency.
- Fastest delivery time windows, identifying optimal delivery periods.
 |
|---|
 |
 |
|---|
- Optimize Load Distribution: Increase dasher availability during peak hours (18:00, 20:00, 22:00) to reduce delays and improve order fulfillment.
- Peak Hour Incentives: Introduce bonuses for dashers during high-demand periods to ensure adequate coverage.
- Enhance Time Estimates: Improve delivery time predictions by integrating real-time traffic, weather, and location data.
- Predictive Analytics: Leverage AI to forecast demand spikes and dynamically allocate resources to reduce delays.
- Buffer Times: Implement controlled buffer times during peak hours to better manage delays and set realistic customer expectations.
- Expand High-Performing Cuisines: Introduce exclusive or premium versions of top-selling cuisines (e.g., Pizza, Mexican, American) to increase customer spend.
- Localize Menus: Tailor food offerings based on regional preferences (e.g., Japanese, Indian, or local specialties) to enhance customer satisfaction.
- Seasonal Menus: Launch limited-time or holiday-themed menu items to boost engagement and sales during specific seasons.
- Prioritize High-Value Orders: Ensure that orders above $50 receive faster processing and delivery during peak hours.
- Accurate Time Estimates: Provide better estimated delivery times for premium orders, ensuring efficient resource allocation.
- Premium Delivery Option: Introduce an optional "Priority Delivery" service for customers willing to pay extra for guaranteed faster delivery.
- Optimize Peak Hour Resources: Utilize predictive models to dynamically adjust dasher availability and restaurant staff levels.
- Promote Off-Peak Ordering: Offer discounts or incentives for orders placed during off-peak hours to balance demand.
- Flexible Delivery Windows: Provide customers with multiple delivery window options to improve satisfaction and reduce peak-hour congestion.
This project aimed to analyze delivery performance, dasher workload, order volumes, and revenue trends across different markets. The insights generated provide actionable recommendations for improving delivery efficiency, optimizing dasher performance, and increasing revenue while maintaining customer satisfaction.
Better dasher load management during peak hours to reduce delivery delays.
Improved time prediction models using AI-driven analytics for better resource planning.
Focusing on high-revenue markets and top cuisines to maximize earnings.
Incentivizing dashers during peak times to enhance service reliability.
- Implement real-time monitoring to dynamically adjust strategies based on evolving demand.
- Encourage dashers to work peak hours through well-structured incentive programs.
- Continuously analyze customer feedback and market trends to refine strategies for long-term growth.
Check the following file for the full project documentation: DoorDash Optimization Project Documentation
All analyses were performed using SQL queries, including:
- Data extraction
- Data cleaning & transformation
- Revenue & dasher analysis
- Delivery performance evaluation
- Advanced insights & KPI comparisons
Queries are available in the DoorDash_Query_Scripts.sql file in this repository.
- Database: MySQL
- Querying & Analysis: SQL
- Visualization: Power BI
- Cloud Services: Google Cloud (For data storage and processing)
- Development Environment: MySQL Workbench, VS Code
- Documentation: Figma, Adobe
