What is 'Data Redundancy', and how can it be minimized in database design?

Official answer

Redundancy is repeated representation of the same fact. It can cause update anomalies when copies disagree. For example, storing a customer's current phone number on every order requires coordinated updates whenever it changes.

Normalize the current customer fact into a keyed customer table and reference it from orders. A join then reads the current value once maintained there. The foreign key enforces the declared reference; it does not copy arbitrary changed attributes into other tables automatically.

First check whether values really are the same fact. A shipping address captured at order time can intentionally differ from today's customer address. Backups, replicas, caches and analytical summaries also duplicate data for different purposes.

Keep an authoritative source and explicit synchronization, freshness and reconciliation rules for justified copies. Use keys and constraints to prevent unintended duplicates, but do not erase distinct historical records merely because some fields match. Choose normalization and measured denormalization around correctness and access needs, not a promise to eliminate every repeated value.