Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 

README.md

SQL PostgreSQL Google BigQuery Data Analysis

📊 SQL & Database Analysis Portfolio

High-impact SQL projects transforming raw marketing and product event data into actionable business strategies. This folder highlights my ability to handle complex data architectures and extract core performance KPIs.


📖 Performance Metrics Dictionary

To ensure data consistency across platforms, the following business logic was implemented in the analytical scripts:

Metric Definition Business Value
ROMI (Value - Spend) / Spend * 100 Measures the direct profitability of ad spend.
Reach Growth (Current Reach - Previous Reach) / Previous Reach Tracks the velocity of audience expansion MoM.
CVR (Funnel) Purchases / Total Sessions Identifies the efficiency of the e-commerce user journey.
Pearson Corr CORR(Engagement, Purchase) Statistically validates if user activity leads to revenue.

🚀 Projects Overview

1️⃣ Omnichannel Marketing Performance Deep-Dive

Tools: PostgreSQL | DBeaver
File: marketing_performance_analysis.sql

Highlights:

  • Gaps & Islands Analysis: Engineered a "Streak" calculation to identify the longest continuous run of active adsets, ensuring high delivery reliability.
  • Spend Volatility: Aggregated multi-source data to identify daily minimum/maximum spend outliers across Google and Facebook.

2️⃣ GA4 E-commerce User Behavior Analysis

Tools: BigQuery | GA4 Event-Model
File: ga4_ecommerce_user_behavior.sql

Highlights:

  • Funnel Diagnostics: Reconstructed a 7-step conversion pipeline to pinpoint exactly where users drop off before checking out.
  • Landing Page Efficiency: Used Regex to parse URL paths and identify the highest converting entry points for new users.

🛠️ Technical Skill Stack

  • Advanced Queries: CTEs, subqueries, and complex multi-table joins.
  • Window Functions: Expert use of LAG(), ROW_NUMBER(), and PARTITION BY.
  • Schema Handling: Proficiency in BigQuery UNNEST for handling nested JSON event parameters.