A practical MySQL data-cleaning project focused on preparing a real-world layoffs dataset for reliable analysis.

This project demonstrates an end-to-end SQL data-cleaning workflow using MySQL. The raw layoffs dataset was inspected, staged, cleaned, standardized, and prepared for downstream analysis.

The project focuses on data quality rather than exploratory analysis, with particular attention to duplicate records, inconsistent text values, date formatting, missing values, and records without meaningful layoff information.

• Create a safe staging environment without modifying the raw dataset.
• Identify and remove duplicate records.
• Standardize inconsistent text values.
• Convert date values from text into a proper SQL DATE type.
• Handle blank and NULL industry values.
• Remove records where both layoff measures are missing.
• Produce a cleaner dataset ready for analysis.

The raw dataset contains 2,361 records and 9 columns:

• company
• location
• industry
• total_laid_off
• percentage_laid_off
• date
• stage
• country
• funds_raised_millions

A staging table was created with the same structure as the original table. The raw data was then copied into the staging table so the original dataset remained unchanged.

ROW_NUMBER() with PARTITION BY was used across the dataset columns to identify duplicate records.

The raw dataset contained 5 duplicate records, which were removed from the staging table.

Several inconsistent text values were cleaned:

• Leading/trailing whitespace was removed from company names using TRIM().
• Industry values beginning with Crypto were standardized to Crypto.
• Trailing periods were removed from country values such as United States..

The original date values were stored as text. They were converted using:

• STR_TO_DATE()
• ALTER TABLE ... MODIFY COLUMN

The final column was converted to the MySQL DATE data type.

Blank industry values were converted to NULL.

A self-JOIN was then used to populate missing industry values when another record for the same company contained a valid industry.

Records where both total_laid_off and percentage_laid_off were NULL were removed because they did not contain meaningful layoff information.

This removed 361 records after duplicate removal.

The temporary row_num column used for duplicate detection was removed from the final staging table.

The final dataset is substantially cleaner and better suited for further SQL analysis, visualization, or dashboard development.

• MySQL
• CREATE TABLE ... LIKE
• INSERT INTO ... SELECT
• Common Table Expressions (CTEs)
• Window functions
• ROW_NUMBER()
• PARTITION BY
• TRIM()
• STR_TO_DATE()
• ALTER TABLE
• UPDATE
• DELETE
• Self-joins
• NULL handling
• Duplicate detection
• Data standardization

sql-data-cleaning-project/
├── data/
│   └── layoffs_raw.csv
├── docs/
│   └── linkedin_project.md
├── sql/
│   └── data_cleaning.sql
├── .gitignore
└── README.md

1. Create a MySQL database.
2. Import the raw layoffs_raw.csv dataset into a table named layoffs.
3. Open sql/data_cleaning.sql in MySQL Workbench.
4. Run the script in sequence.
5. Review layoff_staging2 after the cleaning process.
6. Use the resulting cleaned table for further analysis.

The SQL script is intentionally kept as a separate file so the cleaning process can be reviewed step by step.

This project demonstrates practical skills in:

SQL • MySQL • Data Cleaning • Data Transformation • Data Quality • CTEs • Window Functions • Data Validation • Missing-Value Handling • Duplicate Detection

The cleaned dataset can be used for a follow-up exploratory data analysis project, such as analyzing layoffs by company, industry, country, funding stage, date, and percentage of workforce affected.