When should one consider denormalization in terms of database design? Justify your answer by providing an example (Max: 150 Words).

Denormalization should be considered in database design when performance optimization is a priority. This is because denormalization reduces the need for complex queries and joins, thus improving read performance. For example, in an e-commerce application, if a report frequently needs to display customer details along with their order details, instead of storing customer and order information in separate tables (normalized), you could store the customer's name and address in the order table itself (denormalized). This way, the application can retrieve all necessary information with a single query instead of multiple queries or joins, thereby improving performance. However, it's important to note that denormalization can lead to data redundancy and potential inconsistencies, so it should be used judiciously.