Project Overview
FoodHub, a food aggregator company connecting customers with multiple restaurants, needed to understand their operational performance and customer behavior patterns. With increasing competition in the food delivery market, they required data-driven insights to optimize operations, improve customer satisfaction, and increase revenue.
Business Challenge
- Increasing competition in the food delivery market
- Need to understand customer ordering patterns
- Uncertainty about restaurant partner performance
- Questions about optimal resource allocation
- Need to improve customer satisfaction ratings
Key Question: How can we use data to optimize operations and increase revenue?
Project Goals
- Analyze 1,800+ food delivery orders to identify operational patterns
- Understand customer ordering behavior and preferences
- Evaluate restaurant performance metrics
- Identify key drivers of customer satisfaction
- Provide actionable recommendations for business improvement
Key Metrics
| Metric | Value |
|---|---|
| Orders Analyzed | 1,898 |
| Restaurants | 178 |
| Revenue from Top 20% | 65% |
| Weekend Order Increase | 40% |
| Avg. Delivery Time | 28 minutes |
| Customer Rating Range | 1-5 stars |
Methodology
1. Data Collection & Preparation
Dataset: 1,898 orders containing:
- Order details (ID, date, time, cost)
- Restaurant information (name, cuisine type)
- Delivery metrics (time, distance)
- Customer ratings
Data Cleaning:
- Handled missing values (3% of delivery times)
- Identified and addressed outliers using IQR method
- Validated data consistency across fields
- Created derived features:
- Order value categories (Low, Medium, High)
- Time segments (Morning, Afternoon, Evening, Night)
- Delivery efficiency metrics
- Day of week and weekend flags
2. Exploratory Data Analysis
Univariate Analysis:
- Distribution of order costs, delivery times, and ratings
- Frequency analysis of cuisine types and days of week
- Summary statistics for key metrics
Bivariate Analysis:
- Correlation between delivery time and customer ratings
- Relationship between order value and cuisine type
- Impact of day of week on order volume
- Restaurant performance segmentation
Temporal Analysis:
- Day-of-week ordering patterns
- Time-of-day trends
- Weekend vs. weekday comparisons
3. Statistical Analysis
- Correlation Analysis: Identified relationships between variables
- Segmentation: Grouped restaurants by performance
- Hypothesis Testing: Validated key findings statistically
- Trend Analysis: Identified temporal patterns
Key Findings & Business Impact
Finding 1: Revenue Concentration Risk
📊 Data Insight:
Top 20% of restaurants generate 65% of total revenue, indicating heavy dependence on key partners.
💼 Business Impact:
- High dependency on small number of partners creates business risk
- Potential vulnerability if key restaurants leave platform
- Opportunity to strengthen strategic partnerships
- Need for partner retention strategy
💡 Recommendation:
Launch VIP Restaurant Partnership Program
- Dedicated account management for top performers
- Priority placement in search results
- Co-marketing investment and promotional support
- Volume-based incentive structure
- Quarterly business reviews
📈 Expected Impact:
- Retain high-value partners, reduce churn risk by 30%
- Increase revenue stability
- Strengthen competitive moat
Finding 2: Weekend Demand Surge
📊 Data Insight:
Weekend orders are 40% higher than weekdays, with peak times between 7-9 PM on Friday-Sunday.
💼 Business Impact:
- Delivery delays during peak times lead to customer dissatisfaction
- Current delivery capacity insufficient for weekend demand
- Lower ratings during high-volume periods
- Lost revenue from unfulfilled or delayed orders
💡 Recommendation:
Implement Dynamic Staffing Model
- Increase weekend delivery capacity by 30%
- Implement shift differential pay (15-20% premium) for peak hours
- Partner with gig economy platforms for surge coverage
- Real-time demand forecasting and driver allocation
📈 Expected Impact:
- Reduce average delivery time by 25% on weekends
- Improve weekend customer ratings by 15%
- Capture $200K+ in previously lost revenue
Finding 3: Delivery Time Threshold Effect
📊 Data Insight:
Orders delivered in >35 minutes receive 45% lower ratings compared to deliveries under 25 minutes.
💼 Business Impact:
- Customer churn due to poor delivery experience
- Negative reviews affecting platform reputation
- Lower restaurant ratings impacting their business
- Reduced repeat order rate
💡 Recommendation:
Deploy Delivery Time Management System
- Real-time alerts at 25-minute threshold
- Route optimization algorithm implementation
- “Fast Delivery Guarantee” for restaurants within 3km radius
- Automated compensation policy for late deliveries (discount voucher)
- Driver performance tracking and training
📈 Expected Impact:
- Increase average rating from 3.7 to 4.2 stars (+13.5%)
- Reduce late deliveries by 40%
- Improve customer retention by 20%
Finding 4: Cuisine Profitability Gap
📊 Data Insight:
American cuisine accounts for 35% of orders but Italian restaurants have 23% higher average order value.
💼 Business Impact:
- High volume in American cuisine but lower revenue per order
- Untapped opportunity in high-margin Italian segment
- Suboptimal revenue mix
💡 Recommendation:
Strategic Cuisine Promotion Campaign
- Feature Italian restaurants in premium placement
- “Premium Italian Night” marketing campaign (Thursday-Saturday)
- Bundle deals highlighting higher-margin items
- Targeted push notifications to users with previous Italian orders
- Content marketing (chef interviews, cuisine education)
📈 Expected Impact:
- Increase Italian cuisine orders by 35%
- Boost average order value by 12% over 6 months
- Additional $300K annual revenue
Overall Business Impact Summary
Projected Results After Implementation
| Metric | Current | Projected | Improvement |
|---|---|---|---|
| Average Rating | 3.7 stars | 4.2 stars | +13.5% |
| Weekend Delivery Time | 42 min | 32 min | -24% |
| Average Order Value | $16.50 | $18.50 | +12% |
| Top Restaurant Retention | 78% | 95% | +17% |
| Weekend Capacity | 100% | 130% | +30% |
Financial Impact
Estimated Annual Revenue Increase: $1.2M - $1.8M
Cost of Implementation: ~$200K (technology + staffing)
ROI: 6-9x in first year
Technical Implementation
Data Preparation
# Load and inspect data
import pandas as pd
import numpy as np
df = pd.read_csv('foodhub_data.csv')
# Handle missing values
df['delivery_time'].fillna(df['delivery_time'].median(), inplace=True)
# Create derived features
df['order_category'] = pd.cut(df['cost_of_order'],
bins=[0, 12, 20, float('inf')],
labels=['Low', 'Medium', 'High'])
df['is_weekend'] = df['day_of_week'].isin(['Saturday', 'Sunday'])
# Feature engineering
df['delivery_efficiency'] = df['delivery_time'] / df['distance']
Exploratory Analysis
import matplotlib.pyplot as plt
import seaborn as sns
# Revenue concentration analysis
top_restaurants = df.groupby('restaurant_name')['cost_of_order'].sum() \
.sort_values(ascending=False).head(int(len(df)*0.2))
revenue_concentration = (top_restaurants.sum() / df['cost_of_order'].sum()) * 100
# Delivery time impact on ratings
plt.figure(figsize=(10, 6))
sns.boxplot(x=pd.cut(df['delivery_time'], bins=[0, 25, 35, float('inf')]),
y='rating', data=df)
plt.title('Delivery Time Impact on Customer Ratings')
Statistical Analysis
from scipy import stats
# Correlation analysis
correlation_matrix = df[['delivery_time', 'cost_of_order', 'rating']].corr()
# Hypothesis testing: Weekend vs Weekday orders
weekend_orders = df[df['is_weekend']]['cost_of_order']
weekday_orders = df[~df['is_weekend']]['cost_of_order']
t_stat, p_value = stats.ttest_ind(weekend_orders, weekday_orders)
Visualizations
Distribution Analysis
- Order value distribution by cuisine type
- Delivery time distribution with rating overlay
- Restaurant performance scatter plot
Temporal Patterns
- Order volume by day of week
- Hourly ordering patterns
- Weekend vs. weekday comparison
Relationship Analysis
- Delivery time vs. customer rating correlation
- Order value vs. cuisine type
- Restaurant concentration analysis
Skills Demonstrated
Technical Skills
- Data Manipulation: Pandas, NumPy
- Statistical Analysis: Correlation, hypothesis testing, segmentation
- Data Visualization: Matplotlib, Seaborn
- Feature Engineering: Creating derived metrics
- Python Programming: Clean, documented, reusable code
Analytical Skills
- Exploratory data analysis
- Pattern recognition
- Root cause analysis
- Segmentation and clustering
- Trend identification
Business Skills
- Problem framing and scoping
- Stakeholder communication
- Actionable recommendation development
- ROI calculation
- Strategic thinking
Key Learnings
Technical Growth
- Advanced Pandas techniques for complex data transformations
- Effective visualization storytelling for business audiences
- Statistical methods for validating insights
Business Acumen
- Translating data insights into revenue opportunities
- Understanding food delivery operational metrics
- Balancing customer satisfaction with profitability
Communication
- Presenting technical findings to non-technical stakeholders
- Creating compelling data narratives
- Developing implementable recommendations
Project Deliverables
| Resource | Description |
|---|---|
| 📊 Interactive Analysis (HTML) | Complete exploratory analysis with all visualizations |
| 📓 Jupyter Notebook | Full Python code with detailed comments |
| 📑 Executive Presentation | 12-slide deck with key findings and recommendations |
Relevant Applications
This project demonstrates skills directly applicable to:
✅ Healthcare Analytics: Patient journey analysis, operational optimization
✅ Business Intelligence: Revenue analysis, customer segmentation
✅ Operations Research: Resource allocation, efficiency improvement
✅ Customer Analytics: Satisfaction drivers, retention strategies
Contact
Interested in discussing this project or similar analysis for your organization?