Adding a new column to a production database is not a checkbox in a ticket. It is a live change to a living system. Done right, it brings clarity, speed, and capability. Done wrong, it brings locks, downtime, and incident reports.
Start with the definition. Know exactly what the new column must hold, its type, its constraints, and its default. Every unnecessary NULL is a future bug. Every poorly-chosen type is a future migration.
Assess traffic. Read queries. Understand indexes. If the new column will be queried often, consider adding it to composite indexes or covering indexes. If it will be updated frequently, be aware of write amplification in large tables.
Plan the deployment. Break the change into steps. First, add the column without constraints. Verify the schema change in staging against realistic data volumes. Then backfill in controlled batches to avoid locking. Finally, add constraints and enforce integrity when data is complete.