Storage tiers
Also known as: storage tiering, hot and cold storage, tiered storage, cloud storage tiers
Storage tiers are levels of storage with different speeds and prices, matched to how often data is used. 'Hot' storage holds frequently accessed data on fast drives and costs more. 'Cool' or 'infrequent access' tiers cost less but charge more to retrieve data. 'Archive' tiers, such as Amazon S3 Glacier Deep Archive or Azure Archive, are extremely cheap for storage but can take hours to retrieve from.
While tiering can cut storage costs dramatically, retrieval fees and delays can surprise people during urgent restores.
Storage tiers, explained at three levels
At home
Storage tiers are like where you keep your clothes. Everyday outfits sit in the closet, easy to grab. Winter coats go in the basement, cheaper space but a trip to fetch. Old keepsakes go to a storage unit across town, very cheap but slow and inconvenient to reach. Cloud storage tiers work the same way, charging less the longer you're willing to wait.
At work
Storage tiers are levels of storage with different prices and speeds. Hot storage keeps frequently used data instantly available at the highest price, cool tiers cost less but charge more to read, and archive tiers are cheapest but can take hours to retrieve. Businesses care because moving rarely used data to cheaper tiers can cut storage bills sharply, as long as restore costs are planned.
At home
Storage tiers trade speed for price. Hot storage keeps data on fast drives, ready instantly, at the highest cost. Cool storage is cheaper per gigabyte but charges more each time you read data. Archive storage is cheapest of all, sometimes on tape, but retrieving files can take hours and costs extra. Systems move data down the tiers as it ages and gets used less, so you pay premium prices only for data you actually need quickly.
At work
Cloud tiers, such as Amazon S3 Standard, Infrequent Access and Glacier, or Azure Hot, Cool, Cold and Archive, differ in storage price, retrieval price, retrieval latency and minimum storage duration. Lifecycle rules move objects automatically, for example to cool after 30 days and archive after 180. Deleting or moving data before its minimum duration incurs early deletion fees. Inside data centres, storage systems tier between SSD and hard drives automatically based on access patterns.
At home
Archive storage looks very cheap per month, which makes it tempting for keeping your only copy of photos or backups. The bill arrives when you need everything back at once: retrieval fees, data transfer charges and waiting hours for files to become available can add up quickly. Some services also charge extra for faster restores. In my view, know the restore cost before trusting archive storage with your only copy.
At work
Classify data by access patterns and retention needs before choosing tiers, and set lifecycle policies that match. Watch minimum duration charges, early deletion fees and per-request costs, which can outweigh storage savings for many small files. Model restore costs and times against RTOs, since archive restores can take hours. Test archive retrievals periodically, and tag data with owners so lifecycle decisions are not made blind.
Real-world examples
At home
Thomas archived years of raw video footage to a cheap cloud archive tier, cutting his bill, but noted it would take hours to download if needed.
At work
Lena manages cloud storage for a Toronto architecture firm. Lifecycle rules move project files older than two years to an archive tier, saving thousands annually.
How it compares
| feature | Hot | Cool | Archive |
|---|---|---|---|
| Storage cost | Highest | Lower | Lowest |
| Retrieval time | Instant | Instant | Hours |
| Retrieval fees | None or low | Moderate | Higher |
| Typical data | Active files | Older backups | Long-term records |
Colder tiers are cheaper to keep and costlier, slower to retrieve.
For analysts
In storage cost reviews, I'd map data by how often it's actually used. Many organizations pay hot-storage prices for files nobody has opened in years.
History & origin
Hierarchical storage management, moving older data from expensive disks to cheaper tape, was used on mainframes from the 1970s. As cloud storage grew, Amazon launched Glacier in 2012 for low-cost archiving, followed by infrequent access tiers and S3 Glacier Deep Archive in 2019, priced around $1 US per terabyte per month. Microsoft and Google added comparable tiers. Automatic tiering features, such as Amazon's S3 Intelligent-Tiering in 2018, moved data between tiers based on use. In my view, storage tiers are one of the easiest cloud cost savings, as long as you plan for how and when data comes back.