A new column in a database should be simple, but it never is. Schema changes carry risk. Bad defaults, null constraints, or careless migrations can lock rows, block writes, and take down APIs. The wrong approach means a slow crawl through gigabytes of data while customers see errors.
Plan the migration. Use an explicit migration script. Avoid blanket ALTER TABLE on large datasets during peak hours. In PostgreSQL, add nullable columns without a default to avoid table rewrites. Backfill in batches. In MySQL, evaluate whether your version supports instant DDL to skip copying data. Always test on production-like loads.
Adding a new column also means updating the application layer. Define the column in code and schema. Update queries, indexes, and constraints consciously. Deploy changes in stages. First, create the column. Then deploy code that can read and write to it. Finally, enforce constraints once the data is consistent.