How we test & rank storage

The one metric: real cost per terabyte

Every product on the site is ranked by its real cost per terabyte ($/TB). The calculation is deliberately simple and auditable: we take the product’s current price and divide it by its usable capacity expressed in terabytes. Capacity in gigabytes is converted to terabytes by dividing by 1,000 (capacity GB ÷ 1,000), matching the decimal convention manufacturers print on the box. So a drive listed at 12,000 GB is treated as 12 TB, and its $/TB is simply price ÷ 12. We compute a parallel cost-per-gigabyte figure the same way for small-capacity media where per-TB numbers get unwieldy.

That single number is what makes a 100 GB memory card and an 18 TB tape cartridge comparable on the same axis. You can see it applied across the whole market on the $/TB rankings.

Where the data comes from

Prices and specifications come from retailer and affiliate product listings — the same public listings you would buy from. For each tracked product we record its name, capacity, form factor, interface/technology, condition, stated warranty and current price, then derive $/TB and $/GB from price and capacity. We do not invent prices, fabricate ‘original’ prices to manufacture a discount, or run countdown timers. A ‘deal’ on this site means only one thing: a genuinely low cost per terabyte relative to its category.

Update cadence

The catalog is refreshed daily. Because prices move, the order of any ranking reflects the latest tracked figures each time a page loads, and the cheapest product in a category can change between visits. We note prices as accurate as of the last update; always confirm the live price at the retailer before buying.

How capacity is normalised

Storage is sold with several different capacity conventions, so we normalise carefully:

• Hard drives, SSDs and flash: we use the manufacturer’s decimal display capacity (e.g. a ‘2 TB’ drive = 2,000 GB) for the $/TB math, and preserve the original display-capacity string for the listing so you see exactly what was advertised.
• Tape: LTO cartridges are ranked on native capacity, never the optimistic ‘compressed’ figure that assumes ideal 2.5:1 data. An LTO-9 tape is treated as 18 TB, not its ~45 TB compressed marketing number, because already-compressed media rarely compresses again.
• Multi-packs: where a listing is a pack of discs or tapes, capacity and price reflect the pack so the $/TB is honest.

How condition is tracked

Every product is flagged as new or used, and that flag is visible on every card and filter. The used bucket spans factory-recertified drives (tested and re-warranted by a vendor) and pulled/as-is drives, which carry very different risk. Used enterprise drives frequently top the value rankings, and we treat them as legitimate — but we are explicit about condition so you can weigh the trade-off and keep proper backups. Our drive reliability guide covers how to vet a used disk.

Ranking philosophy

Default order is always lowest $/TB first, layered with editorial judgement about fitness for purpose. We will never reorder a list because a retailer pays a higher commission — the sort is computed from price and capacity before commissions are even known. Where a cheapest-per-TB option carries a real caveat (an SMR drive in a RAID context, a drive that needs a SAS HBA, flash left unpowered for years), we say so in the surrounding guidance rather than silently burying or boosting it.

What we don’t do

• We don’t accept payment to change rankings, and commissions never affect order. (See the affiliate disclosure.)
• We don’t show fake discounts, invented list prices or artificial scarcity.
• We don’t run lab endurance or teardown testing — we are a price and value tracker, not a hardware test bench, and we don’t pretend otherwise.
• We don’t hide the condition of a drive or quote compressed tape capacities as if they were real.

For the reasoning behind the whole approach, and who builds it, see about harddisc.net.