Understanding Database Normalization in SQL with Example

See how raw, repetitive data gets organized into clean, related tables.

Normalization is the process of structuring a database to reduce redundancy and prevent data inconsistencies. Rather than explaining it in the abstract, it's easiest to understand through an example table that gets progressively cleaned up.

The Unnormalized Starting Point

OrderID | Customer | Product1  | Product2
1       | Riya     | Pen       | Notebook
2       | Aman     | Pen       | NULL

Storing multiple products in separate columns makes it hard to add a third product, and wastes space with NULLs.

First Normal Form (1NF)

Eliminate repeating groups — each row should hold a single value per column.

OrderID | Customer | Product
1       | Riya     | Pen
1       | Riya     | Notebook
2       | Aman     | Pen

Second Normal Form (2NF)

Remove partial dependency — every non-key column must depend on the whole primary key, not just part of it. This usually means splitting into separate Orders and OrderItems tables.

Orders: OrderID, Customer
OrderItems: OrderID, Product

Third Normal Form (3NF)

Remove transitive dependency — non-key columns shouldn't depend on other non-key columns. If Customer also stored a CustomerCity, that belongs in its own Customers table instead.

Customers: CustomerID, Name, City
Orders: OrderID, CustomerID
OrderItems: OrderID, Product

💡 Tip: Most production databases stop at 3NF — going further (BCNF, 4NF) is usually reserved for very specific edge cases.

Why Normalize?