The query runs, but the results are wrong. You check the data model. The reason is obvious: the schema changed, but the code did not. A new column exists in the database, yet the application does not know it.
Adding a new column is simple in theory. In practice, it can break everything if done without control. The process always starts with a schema migration. Define the column name, type, and defaults. Keep the migration file in version control so every environment stays in sync.
After the migration, update the ORM models or data access layer. If the new column is nullable, ensure new writes handle default values. If it is required, backfill historical rows before enforcing constraints. Do not skip data audits—missing values create silent bugs that surface later.
Test the changes locally. Run integration tests against a clone of production data. Confirm read and write operations for the new column behave as expected. Watch for query regressions, especially if the column is indexed. Adding indexes speeds reads but can slow inserts or updates. Benchmark before deploying.