Global Tech Layoffs Trend Analysis

SQL Data Cleaning + Exploratory Data Analysis

+ Tableau Dashboard

---

Project Overview

An end-to-end data analysis project examining global tech industry layoffs across companies, industries, countries and funding stages. Raw messy data was cleaned using SQL in DB Browser for SQLite, analysed through structured EDA queries, and visualised in an interactive Tableau Public dashboard.

---

🔗 Live Dashboard

[View on Tableau Public]https://public.tableau.com/app/profile/ankita.s3430/viz/GlobalLayoffsTrendAnalysis/Dashboard1

---

Dataset

• Source: Kaggle — World Layoffs Dataset
• Rows: 2,361 records
• File: layoffs.csv → layoffs_cleaned.csv
• Coverage: Global tech layoffs 2020–2024

Columns:

Column	Description
company	Company name
location	Office location
industry	Industry sector
total_laid_off	Total employees laid off
percentage_laid_off	% of workforce laid off
date	Date of layoff event
stage	Company funding stage
country	Country of layoff
funds_raised_millions	Total funds raised (USD M)

---

Tools Used

• DB Browser for SQLite — Data cleaning & EDA
• Tableau Public — Dashboard & visualisation
• GitHub — Version control & documentation

---

Data Cleaning (SQL)

1. NULL Value Check

SELECT
COUNT(*) - COUNT(total_laid_off) 
    as null_total_laid_off,
COUNT(*) - COUNT(percentage_laid_off) 
    as null_percentage_laid_off,
COUNT(*) - COUNT(industry) 
    as null_industry,
COUNT(*) - COUNT(funds_raised_millions) 
    as null_funds_raised,
COUNT(*) - COUNT(stage) 
    as null_stage,
COUNT(*) - COUNT(country) 
    as null_country,
COUNT(*) - COUNT(date) 
    as null_date
FROM layoffs;

2. Duplicate Detection

SELECT company, location, industry,
       total_laid_off, percentage_laid_off,
       date, stage, country, 
       funds_raised_millions,
       COUNT(*) as duplicate_count
FROM layoffs
GROUP BY company, location, industry,
         total_laid_off, percentage_laid_off,
         date, stage, country,
         funds_raised_millions
HAVING COUNT(*) > 1;

3. Removing Duplicates Using ROW_NUMBER()

CREATE TABLE layoffs_cleaned AS
WITH duplicate_cte AS (
    SELECT *,
           ROW_NUMBER() OVER(
               PARTITION BY company, location,
               industry, total_laid_off,
               percentage_laid_off, date,
               stage, country,
               funds_raised_millions
               ORDER BY company
           ) as row_num
    FROM layoffs
)
SELECT * FROM duplicate_cte
WHERE row_num = 1;

4. Standardising Company Names

UPDATE layoffs_cleaned
SET company = TRIM(company);

5. Filling Missing Industry Values

UPDATE layoffs_cleaned t1
SET industry = (
    SELECT industry
    FROM layoffs_cleaned t2
    WHERE t1.company = t2.company
    AND t2.industry IS NOT NULL
    LIMIT 1
)
WHERE t1.industry IS NULL;

6. Standardising Industry Values

UPDATE layoffs_cleaned
SET industry = 'Crypto'
WHERE industry LIKE 'Crypto%';

Cleaning Summary

Check	Finding	Action Taken
Duplicates	Found across multiple columns	Removed using ROW_NUMBER() CTE
NULL Industries	Several records missing	Filled using same-company records
Company Names	Leading/trailing spaces	Trimmed using TRIM()
Industry Values	Crypto variations found	Standardised to single value
Country Names	Checked for inconsistencies	Verified and confirmed clean

---

Exploratory Data Analysis (SQL)

Maximum Layoffs in Single Event

SELECT MAX(CAST(total_laid_off AS INT)) 
    AS max_layoffs
FROM layoffs_cleaned;

Top 10 Companies by Total Layoffs

SELECT company,
       SUM(CAST(total_laid_off AS INT)) 
           AS total_layoffs
FROM layoffs_cleaned
GROUP BY company
ORDER BY total_layoffs DESC
LIMIT 10;

Industries with Highest Layoffs

SELECT industry,
       SUM(CAST(total_laid_off AS INT)) 
           AS total_layoffs
FROM layoffs_cleaned
GROUP BY industry
ORDER BY total_layoffs DESC;

Countries with Highest Layoffs

SELECT country,
       SUM(CAST(total_laid_off AS INT)) 
           AS total_layoffs
FROM layoffs_cleaned
GROUP BY country
ORDER BY total_layoffs DESC;

Layoffs by Year

SELECT SUBSTR(date, 1, 4) AS year,
       SUM(CAST(total_laid_off AS INT)) 
           AS total_layoffs
FROM layoffs_cleaned
GROUP BY year
ORDER BY year;

Rolling Monthly Total (Window Function)

WITH monthly_layoffs AS (
    SELECT SUBSTR(date,1,7) AS month,
           SUM(CAST(total_laid_off AS INT)) 
               AS total_layoffs
    FROM layoffs_cleaned
    GROUP BY month
)
SELECT month, total_layoffs,
       SUM(total_layoffs) OVER (
           ORDER BY month
       ) AS rolling_total
FROM monthly_layoffs;

Companies with Multiple Layoff Events

SELECT company,
       COUNT(*) AS layoff_events
FROM layoffs_cleaned
GROUP BY company
HAVING layoff_events > 1
ORDER BY layoff_events DESC;

Layoffs by Funding Stage

SELECT stage,
       SUM(CAST(total_laid_off AS INT)) 
           AS total_layoffs
FROM layoffs_cleaned
GROUP BY stage
ORDER BY total_layoffs DESC;

---

Key Business Insights

1. US Dominates Global Layoffs — United States accounts for the largest share of total layoffs globally, reflecting concentration of major tech employers

2. Consumer and Retail Industries Hit Hardest — Consumer-facing and retail tech sectors recorded highest total layoffs indicating post-pandemic demand correction

3. Peak Layoff Period: 2022-2023 — Layoff events surged significantly following the post-pandemic tech hiring boom, with rolling monthly totals accelerating sharply

4. Large Funded Companies Not Immune — Several companies with hundreds of millions in funding still conducted significant layoff rounds, disproving the assumption that capital protects against workforce reduction

5. Post-IPO Stage Companies Hit Hardest — Analysis by funding stage reveals Post-IPO companies recorded the highest total layoffs, suggesting pressure from public market performance expectations

6. Companies with Repeated Layoff Events — Multiple companies conducted more than one round of layoffs, indicating structural rather than one-time workforce adjustments

7. Crypto Industry Consolidation — Crypto sector showed significant layoffs consistent with the broader 2022 crypto market collapse

---

Recommendations

• Companies with multiple layoff events should be monitored as indicators of deeper structural instability
• Geographic diversification of tech workforce reduces concentration risk
• High funding alone is insufficient predictor of workforce stability
• Seasonal and macroeconomic triggers should inform hiring cycle planning

---

Dashboard Preview

![Dashboard]

---

Project Structure

global-layoffs-trend-analysis/ │ ├── data/ │ ├── layoffs.csv (raw) │ └── layoffs_cleaned.csv ├── sql/ │ ├── data_cleaning.sql │ └── eda_queries.sql ├── dashboard/ │ └── GlobalLayoffsTrendAnalysis.twbx ├── screenshots/ │ └── dashboard.png └── README.md