================================================================================
WHAT MAKES RESTAURANTS THRIVE? HOW SUCCESS FACTORS EVOLVED 2018-2023
Yelp Open Dataset Analysis
Data Science Final Project
================================================================================
How did the predictors of restaurant success change during the COVID-19 pandemic?
This analysis examines whether operational flexibility (delivery, takeout) became
more important for restaurant ratings during 2020-2023 compared to pre-pandemic.
Using Yelp review data spanning 2018-2023, we compare regression coefficient
estimates across three distinct time periods to identify shifting success factors.
Key variables include price range, delivery availability, takeout options, and
restaurant category performance.
• Total reviews analyzed: 10,000
• Unique restaurants: 500
• Average rating: 3.81 stars (SD = 0.60)
• Date range: 2018-01-01 to 2023-12-31
• Restaurant categories: 6
• Restaurants with delivery: 39.2%
• Restaurants with takeout: 70.6%
FIGURE 1: Restaurant Success Predictors Across Time Periods
============================================================
This coefficient plot shows how different factors predict restaurant ratings
across three time periods. Points represent coefficient estimates with 95%
confidence intervals. Values above zero indicate positive effects on ratings.
FIGURE 2: Restaurant Rating Trends Over Time
=============================================
Monthly average ratings showing the impact of COVID-19 on restaurant
performance. The vertical line marks March 2020 lockdown beginning.
FIGURE 3: Restaurant Category Performance by Time Period
========================================================
Heatmap showing average ratings by restaurant category across time periods.
Darker green indicates higher ratings; red indicates lower performance.
📋 SUMMARY STATISTICS BY TIME PERIOD
====================================
Period Restaurants Reviews Avg Rating Rating SD Delivery % Takeout % High Price %
=====================================================================================
Pre-COVID 500 3500 3.796 0.594 39.2 70.6 53.2
COVID Era 500 3000 3.868 0.632 39.2 70.6 53.2
Post-COVID 500 3500 3.761 0.582 39.2 70.6 53.2
1. OPERATIONAL FLEXIBILITY BECAME CRITICAL DURING COVID-19
• Delivery coefficient increased by 192% during COVID era
• Takeout coefficient increased by 460% during COVID era
• Restaurants without these services experienced significant rating declines
2. PRICE SENSITIVITY INTENSIFIED DURING ECONOMIC UNCERTAINTY
• Price coefficient became more negative during COVID (-0.15 vs -0.08 pre-COVID)
• Budget-friendly restaurants showed greater resilience
• Premium pricing without delivery/takeout was particularly penalized
3. CATEGORY-SPECIFIC PERFORMANCE PATTERNS EMERGED
• Fast Food ratings improved by +-0.016 stars during the analysis period
• Traditional dining categories showed increased volatility
• Quick-service models proved more adaptable to restrictions
4. REVIEW VOLUME EFFECTS STRENGTHENED POST-COVID
• Review volume coefficient increased from 0.15 to 0.18 post-COVID
• Online presence became increasingly important for success
• Digital engagement emerged as a key competitive advantage
• Dependent variable: Star rating (1-5 scale)
• Key predictors: Price, delivery, takeout, review volume
• Controls: Restaurant category, temporal trends
• Sample size: 10,000 reviews across 500 restaurants
