The jargon behind our test benches and spec tables, defined in plain language.
A door lock's independently tested attack-resistance rating (Grade 1 to 3, Grade 1 being the most secure), covering things like forced-entry and lock-picking resistance.
The usable food capacity of an air fryer's basket — distinct from the manufacturer's advertised total capacity, which often includes space you can't actually load and still get even airflow.
Whether a spec came from our own bench test (measured) or from the manufacturer's own materials (claimed). See our Methodology page for which parameters are allowed to be claimed-only.
The total real cost of a product or plan over a stated period (usually 1-3 years), including renewal price hikes, add-on fees, and anything else the sticker price doesn't show.
The layered midsole materials inside a shoe (foam type, density, insert), which determine ride feel more than any single spec number does.
Whether a lock defaults to unlocked (fail-safe) or locked (fail-secure) when it loses power or battery — a real safety trade-off, not just a spec-sheet line.
For slip-on/hands-free shoes, how often a shoe can be put on without using your hands, tracked across repeated attempts during testing.
Indentation Load Deflection — the force needed to compress a foam layer by 25% of its height. It's the actual engineering measure of foam firmness, independent of a brand's own "soft/medium/firm" labeling.
The discounted first-term price shown in most hosting/subscription marketing — not what you'll pay on renewal.
How often a smart lock's keypad correctly reads a code on the first try, tracked over a set number of real entry attempts during testing.
How much movement on one side of a mattress is felt on the other side — measured by dropping a weighted ball on one side and reading vibration on the other. Lower is better if you share a bed.
How long an appliance takes to reach its target temperature from cold, timed on our own bench rather than taken from the manual.
What you actually pay after an introductory period ends — often several times the advertised "intro" rate. We track this separately because it's the number that determines real cost of ownership.
Time To First Byte — how long a server takes to send the first byte of a response after a request. The clearest single number for "how fast does this host actually respond," measured independently rather than taken from the host's own dashboard.
A surface-temperature measurement taken after sustained contact, used to score how hot or cool a mattress or cushion actually runs — not a marketing claim about "cooling technology."
Every measured comparison parameter has a tolerance — a gap small enough to be measurement noise rather than a real difference. Two products inside that tolerance are marked a tie instead of one being crowned a winner.
The uptime percentage a host contractually guarantees, and what credit (if any) you're owed if they miss it. A 99.9% SLA still allows for about 8.7 hours of downtime a year.
A real interview with a confirmed product owner, conducted directly by our team — not a scraped or aggregated online review.
Shoptera's rotating group of testers who put footwear and wearables through real daily use over weeks, rather than a single unboxing session — the source of our comfort and durability notes.
The specific combination of sub-scores used to compute an overall rank for a particular use case — "Best for families" and "Best value" can rank the same products differently because they weight sub-scores differently.