A review is only worth reading if you know what produced it. This page describes how teqpost arrives at a verdict, so you can judge whether our testing resembles your use.
We use the thing
Hardware is used as a primary device for a sustained period before we publish — not benchmarked for an afternoon. Most of what matters about a laptop or a phone (fan behaviour, hinge wear, battery decay, software irritations) only surfaces after the novelty has gone.
Sustained load, not peak numbers
Where we quote performance, we report what a device sustains under prolonged load rather than the highest figure it reaches in the first thirty seconds. Peak numbers are a property of a cold heatsink and tell you very little. The reasoning is set out in how to read a laptop spec sheet.
The conditions are stated
- How long we used the device, and as what.
- What we ran, at what settings, for how long.
- Screen brightness and power profile, because both dominate battery results.
- Firmware and operating system versions at time of testing.
Who a product is wrong for
Every review names the reader who should not buy the product. A verdict that recommends something to everyone has not been thought about.
Loans, samples and purchases
Where a manufacturer supplies a review unit, we say so in the article. Review units are returned or disclosed as retained. Manufacturers do not see copy before publication and have no approval over it.
Scores
We do not publish a single aggregate score. A number implies a precision we cannot support and flattens the tradeoffs that actually decide whether a product suits you.
Corrections
Factual errors are corrected in place with a note describing what changed and when. If a correction materially alters a verdict, that is stated at the top of the article. To report one, contact us.