This project was developed as part of the MIT Professional Education - Applied AI & Data Science Program, a comprehensive diploma program focused on practical applications of artificial intelligence and data science in business contexts.
The analysis focuses on FoodHub, a New York-based food aggregator company. The goal is to analyze order data to understand restaurant demand, evaluate operational performance (preparation and delivery times), measure customer satisfaction, and provide data-driven recommendations to enhance business operations and customer experience.
The project demonstrates end-to-end data science skills, including data cleaning, exploratory data analysis (EDA), statistical analysis, and generating actionable business insights.
- Demand Analysis: Identify top-performing restaurants and cuisine types, distinguishing between weekday and weekend patterns.
- Operational Evaluation: Calculate and analyze preparation and delivery times, identifying potential bottlenecks.
- Customer Satisfaction: Explore relationships between customer ratings and factors such as order cost and delivery time.
- Business Insights: Generate concrete recommendations to improve customer experience, optimize operations, and maximize revenue.
The analysis uses foodhub_order.csv, containing 1,898 orders with the following attributes:
| Column | Description |
|---|---|
order_id |
Unique order identifier |
customer_id |
Customer identifier |
restaurant_name |
Name of the restaurant |
cuisine_type |
Type of cuisine ordered |
cost |
Order cost (USD) |
day_of_the_week |
Weekday or Weekend |
rating |
Customer rating (1-5) or "Not given" |
food_preparation_time |
Preparation time (minutes) |
delivery_time |
Delivery time (minutes) |
Shake Shack leads with 219 orders, followed by The Meatball Shop and Blue Ribbon Sushi. A small cluster of restaurants drives the majority of platform volume.
American cuisine dominates across both periods, with 415 weekend orders. Weekend demand spikes consistently across all top categories, confirming peak-period concentration.
The majority of orders fall in the $10–$20 range. ~29% exceed the $20 threshold, triggering the 25% commission tier and representing a disproportionately valuable segment for revenue.
39% of orders (736) have no rating submitted. Among rated orders, 4 and 5-star reviews dominate, suggesting high satisfaction — though the large unrated pool limits conclusive analysis.
Weekday deliveries average 28.34 minutes vs. 22.47 on weekends — a ~6-minute gap likely driven by traffic and courier supply. ~10.5% of all orders exceed 60 total minutes (prep + delivery).
- Language: Python 3.8+
- Environment: Google Colab / Jupyter Notebook
- Data Manipulation:
pandas,numpy - Visualization:
matplotlib,seaborn - Statistics: Built-in Python statistics libraries
The analysis revealed several critical insights:
- Weekend-Driven Demand: Orders are significantly higher on weekends, representing peak demand periods.
- Cuisine Concentration: American cuisine dominates, with 415 weekend orders, indicating clear core categories.
- Restaurant Concentration: A small group of restaurants drives most volume. Shake Shack leads with 219 orders, followed by The Meatball Shop (132), Blue Ribbon Sushi (119), Blue Ribbon Fried Chicken (96), and Parm (68).
- Operational Performance:
- Average preparation time: 27.37 minutes
- Average delivery time: 24.16 minutes
- ~10.54% of orders exceed 60 minutes total (preparation + delivery)
- Weekday deliveries are ~6 minutes slower than weekends (28.34 min vs. 22.47 min)
- Customer Satisfaction:
- 736 orders (39%) have no rating, limiting satisfaction analysis
- Rated orders show predominantly positive feedback (4-5 stars)
- Near-zero correlation between ratings and cost/delivery time suggests other factors (food quality, accuracy) drive satisfaction
- Revenue Structure: Under the tiered commission model (25% for orders >$20, 15% for orders >$5), total net revenue is $6,166.30. Approximately 29.24% of orders exceed $20, representing a valuable premium segment.
- Promotion-Eligible Restaurants: Restaurants meeting criteria (>50 ratings, avg rating >4) are:
- The Meatball Shop (avg rating: 4.51)
- Blue Ribbon Fried Chicken
- Shake Shack (133 ratings)
- Blue Ribbon Sushi
- Weekend-focused campaigns: Concentrate major promotions and featured placements on weekends to maximize conversion.
- Price-based segmentation: Offer bundle deals for the value segment (<$20) to increase basket size, while curating "Premium Picks" collections for the >$20 segment.
- Cuisine visibility: Prioritize top cuisines (American, Japanese) in discovery features while rotating low-volume cuisines to test demand growth.
- Support key partners: Provide operational support to high-volume restaurants (e.g., Shake Shack) and explore co-marketing agreements.
- Data-driven promotions: Use the established criteria (>50 ratings, avg rating >4) to select restaurants for ad placements and "Recommended" badges, ensuring quality and reliability.
- "Rising Stars" program: Give controlled visibility to high-rated but low-sample restaurants to diversify supply without compromising customer experience.
- Target 60+ minute orders: Analyze root causes (preparation vs. delivery delays) and implement corrective actions with restaurants and couriers.
- Improve weekday delivery: Increase courier supply during weekday peak windows (lunch/dinner) and optimize routing to reduce the 6-minute gap compared to weekends.
- Reduce unrated orders: Implement post-delivery nudges (one-tap rating prompts) or small incentives (points, discounts) to increase rating response rates, enabling more reliable satisfaction analysis.
Gabriela Yasmin Vidales Ayala
Applied AI & Data Science Program
MIT Professional Education
This project was developed as a capstone assignment for the MIT Professional Education diploma program, demonstrating applied data science skills in a real-world business context.
This project is licensed under the MIT License. See the LICENSE file for details.