Amazon S3 offers a range of storage classes designed for different use cases. The Amazon S3 storage classes are listed here:

• Amazon S3 Standard: For general purpose storage of frequently accessed data
• Amazon S3 Standard - Infrequent Access: For long-lived, but less frequently accessed data
• Amazon Glacier: For long-term archive

Taking advantage of Amazon S3 - Infrequent Access (IA) instead of Standard S3 can lead to immediate storage costs savings up to 30% with no code changes. You will need to understand your storage usage patterns and then configure policies to automatically move all or substantial portion of your data to the lower cost storage class, appropriately. 

You can use S3 "Static" Website Hosting to even eliminate your web server tier. Static in quotes because it is not exactly static in nature – you can do a bunch of active stuff as it does support JavaScript (includes AWS SDK) and Cross-Origin Resource Sharing (CORS). The advantages include no servers, no patching, and no scaling rules required. Given that web page sizes are increasing, this approach can avoid costs associated with web server patching, capacity planning, and security scanning, while making content rollouts/updates easier (less testing required).

Amazon Glacier storage class can be used to store backups and archive old data. Amazon Glacier is low-cost storage with 99.999999999% data durability. Data restores from Glacier storage can take anywhere from a few minutes to a few hours. However, this can result in 50 to 60% savings on storage. You can also specify life cycle rules to automate data movement from S3 to Glacier storage.

Using the Reduced Redundancy Storage (RRS) option in Amazon S3 storage can reduce your costs by storing noncritical and easily restorable data at lower levels of redundancy than the standard storage option. Amazon S3’s reduced redundancy option stores data in multiple facilities and on multiple devices, but the data is replicated fewer times. RRS provides 99.99% data durability versus 99.999999999% using the standard option. This can lead to savings of additional 15 to 20% on storage.

Databases (particularly RDBMSs) make poor BLOB stores and are a poor choice in terms of performance, management, and costs especially Multi-AZ settings. Store blobs on S3 with reference URLs in DB.