🗄️ Global Layoffs SQL Analysis

Exploring worldwide tech layoff trends through data cleaning, transformation, and exploratory analysis — all in SQL.

---

📌 Overview

Between 2020 and 2024, thousands of companies worldwide laid off hundreds of thousands of workers. This project digs into that data using SQL to answer key questions:

• Which industries and companies were hit hardest?
• How did layoffs evolve over time?
• What geographic regions saw the most impact?
• Which funding stages were most affected?

The focus is on demonstrating real-world SQL skills — from messy raw data to clean, insightful analysis.

---

🎯 Project Goals

• Practice data cleaning and transformation in SQL
• Perform exploratory data analysis (EDA) without external tools
• Uncover meaningful patterns in a real-world dataset
• Build a portfolio piece that showcases end-to-end SQL thinking

---

🗂️ Project Structure

Layoffs_Project_SQL/
│
├── data_cleaning.sql        # Raw → clean: deduplication, nulls, formatting
├── exploratory_analysis.sql # EDA queries and trend analysis
├── data/
│   └── layoffs.csv          # Raw dataset
└── README.md

---

🧹 Part 1 — Data Cleaning

Raw data is rarely analysis-ready. This stage covers:

• ✅ Removing duplicate records
• ✅ Standardising inconsistent values (company names, industries, countries)
• ✅ Handling NULL and blank values
• ✅ Fixing date formats for time-series analysis
• ✅ Dropping irrelevant or unreliable columns

---

📊 Part 2 — Exploratory Data Analysis

Key questions explored through SQL queries:

Question	Approach
Which companies had the largest single layoffs?	ORDER BY total_laid_off DESC
Which industries were hit hardest overall?	GROUP BY industry
How did layoffs trend month over month?	Rolling sum with OVER() window functions
Which countries saw the most layoffs?	GROUP BY country
Which funding stages were most affected?	GROUP BY stage
Top 5 companies by layoffs per year	DENSE_RANK() window function

---

💡 Key Findings

• Consumer and Retail industries saw some of the highest total layoffs
• Layoff peaks aligned with major market downturns and post-pandemic corrections
• Several companies laid off 100% of their workforce, signalling full shutdowns
• Later-stage funded companies (Post-IPO) accounted for a large share of total layoffs
• Layoffs were heavily concentrated in the United States

---

🛠️ SQL Concepts Used

Concept	Used For
CTEs	Multi-step cleaning logic
ROW_NUMBER()	Deduplication
DENSE_RANK()	Yearly company rankings
Window functions OVER()	Rolling totals over time
GROUP BY + ORDER BY	Aggregated trend analysis
CASE WHEN	Conditional transformations
String functions	Standardising messy text fields

---

🚀 How to Run

1. Clone the repo

git clone https://github.com/gail-mar/Layoffs_Project_SQL.git
cd Layoffs_Project_SQL

2. Set up the database

Import the dataset into your SQL environment (MySQL recommended):

CREATE DATABASE layoffs_db;
USE layoffs_db;
-- then import layoffs.csv via your SQL client

3. Run the scripts in order

1. data_cleaning.sql
2. exploratory_analysis.sql

---

📁 Dataset

The dataset contains real-world layoff records including:

• Company name, industry, and location
• Total employees laid off & percentage laid off
• Date of layoffs
• Company stage and funds raised

Source: Layoffs.fyi — a crowd-sourced tracker of tech layoffs.

---

👩‍💻 Author

Gail Marechal — Data Science & AI

---

Part of a data analytics portfolio showcasing SQL, Power BI, and AI projects.