How we test

How we know what we know

Before you trust a single pick, you deserve to know how we reached it. Every recommendation on this site carries a label that tells you exactly that, and we use three of them — never blurring the line between one and the next. That label is why a scoped claim like "the best chef's knife for small hands" is something you can check rather than take on faith: it points straight to the evidence behind the words.

The three evidence tiers

We tested this

Our highest tier. It means a member of the editorial team put the product to hands-on use, followed a written test protocol, and recorded original notes, photos, or measurements along with the date we tested it. When you see this label, the judgments come from our own hands, not a spec sheet. We tell you what we measured and how, so you can weigh it for yourself.

We researched this

This tier means we did not use the product hands-on, but we built a structured assessment from authentic manufacturer specifications and a defined set of credible third-party, expert, and verified-owner sources, judged against written criteria. It is a real, careful synthesis — and we say plainly that it is research, not first-hand use, so you never mistake one for the other. We also state what we did not do.

We synthesized this

It means we formed an editorial judgment from aggregated public information — manufacturer specifications, documented design, and each product's established reputation — rather than from our own hands-on test or a deeper source-based study. It is the tier behind our recommendations today, as we build hands-on testing toward the stronger tiers above. We never hide it: every synthesised pick states its basis and its limits in plain words and names who it is wrong for, so you can weigh the evidence yourself rather than take on faith a test we have not run.

Every pick names its basis — and who it is wrong for

Two rules keep the tiers honest. First, every pick shows its tier and the basis for it, so you always know the strength of the evidence. Second, every pick names the reader it is wrong for. A recommendation that suits everyone suits no one; naming the misfit is our plainest proof that we are assessing a product, not just selling it.

What we never do

We invent no data — not a rating, not a review count, not a star total, not a test we did not run, not a spec we cannot trace to an authentic source. If we do not know something, we say so, or we leave it out. An absent fact is acceptable; a fabricated one would disqualify the whole page. It is also why we do not publish numbers that go stale — we point you to Amazon for the current figure instead.

Why this matters to you

Labels turn a claim into something you can check. When we call a pick "the best for" a specific job, that superlative is scoped to a reader and backed by the tier shown beside it. You never have to take our word on faith — you can see the basis, weigh it against your own needs, and decide. That is the whole point. For how we stay independent while doing this, see our editorial policy.