Kaggle Zomato Data Analytics Project

Overview

A full analytics workflow from raw Kaggle data to curated SQL views and an interactive Power BI dashboard for restaurant market intelligence.

Problem Statement

The dataset contained missing values, duplicate records, and inconsistent cuisine/location tags that blocked reliable market comparisons.

Dataset

Kaggle Zomato restaurant dataset with location, cuisine, rating, price, and delivery attributes

Data Cleaning

• Handled missing ratings and price values using business rules
• Removed duplicate restaurant records after fuzzy name matching
• Standardized cuisine tags and city naming conventions
• Created cleaned staging tables before loading analysis-ready views

SQL Queries

SELECT
  City,
  COUNT(*) AS RestaurantCount,
  AVG(Rating) AS AvgRating,
  AVG(Votes) AS AvgVotes
FROM dbo.Restaurants_Clean
GROUP BY City
ORDER BY AvgRating DESC;

SELECT
  Cuisine,
  COUNT(*) AS Outlets,
  AVG(AggregatedRating) AS AvgRating,
  SUM(CASE WHEN HasOnlineDelivery = 1 THEN 1 ELSE 0 END) AS DeliveryEnabled
FROM dbo.Restaurants_Clean
GROUP BY Cuisine;

Python Analysis

• Performed EDA on rating distributions and price bands
• Identified cuisine clusters with high vote volume and delivery adoption

Power BI Dashboard

Interactive dashboard with slicers for city, cuisine, price range, and delivery availability. KPI cards for average rating, vote volume, and online delivery share with segment comparison visuals.

Business Insights

• Certain cuisines showed consistently higher ratings and order potential in target cities
• Mid-price restaurants had the strongest balance of volume and satisfaction
• Delivery-enabled outlets correlated with higher customer engagement in urban markets

Recommendations

• Target expansion in cuisines with high demand but lower outlet density
• Prioritize delivery partnerships in cities with adoption gaps
• Use rating and vote thresholds to shortlist high-confidence market entries

Technologies