The query hit production at 14:03, and the reports were wrong. A missing field. A schema out of sync. You need a new column. Now.
Adding a new column should not break a system, but done poorly, it will. Every database engine supports altering a table to add a column, but the operational reality depends on size, load, and lock behavior. On small tables, it can be instantaneous. On large tables under heavy writes, it can stall queries, consume I/O, and block critical paths.
Plan before you run the ALTER TABLE. Check the engine documentation. Some versions of MySQL and PostgreSQL can add a nullable column with a default without a full table rewrite. Others cannot. For large datasets, use an online schema migration tool like pt-online-schema-change or gh-ost. These gradually create a new table with the desired structure and swap it in, avoiding downtime.
In analytics pipelines, adding a column means updating data models, ETL jobs, and dashboards. Failing to propagate the new column through every dependent system leads to silent failures or broken views. Treat the change as schema evolution, not a single command. Track versioned migrations in code, commit them to source control, and review them like application changes.